National Academies Press: OpenBook
« Previous: The Navigation Feasability Study
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

5

Improving Waterway System Planning

Like many large water resources project planning studies, the Upper Mississippi River–Illinois Waterway system navigation feasibility study represents a complicated challenge. Even with a large budget, decades to complete the study, and an environment free of the contending interest groups, the conclusions would contain a significant degree of uncertainty. The committee recognizes the inherent difficulty of the Corps' task and our inability to give precise answers to the questions that the Corps must answer.

A major difficulty is the conflict between long-term infrastructure investments and the rapidly changing nature of the U.S. economy. Since the Second World War, the role of U.S. agriculture has changed from feeding the world through massive surplus production to exporting into a highly competitive world market for food and fiber. Increasing world population has increased food demand, but the Green Revolution and other innovations in agronomy have increased food supply throughout the world. American agricultural exports compete with exports from Argentina, Brazil, and other food exporters, as well with increasing domestic production in importing nations. The dynamic nature of the competition and changes in technology make it difficult to accurately forecast future levels of American food exports.

Infrastructure improvements, such as larger locks on the Upper Mississippi River, are designed to expedite river traffic for the next century or more. In evaluating the social benefits of these improvements, the benefits 20–100 years into the future are important. Unfortunately, rapid changes in the U.S. economy and in other national economies mean that the future value of these navigation improvements is uncertain. Growing world population, growing affluence, and innovations in agriculture could lead to either much greater or much lower demand for U.S. grain exports. Extending large locks on the Upper Mississippi could be similar to the example of the construction of the huge airport in Gander, Newfoundland, which was built to serve propeller aircraft that needed to stop on transatlantic flights. Rapidly increasing traffic at the airport in the early 1950s induced the authority to make an expensive infrastructure investment just as jets were being introduced that would fly nonstop between Europe and the U.S., thereby avoiding Gander.

A rapidly changing economy makes the future benefits of infrastructure investment extremely uncertain: however, this cannot and should not be a reason to delay or stop investment.

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

If it was decided to not invest in lock extensions, traffic on the waterway would likely not grow much beyond current levels because of the costs associated with congestion. Thus, although uncertainty about the future value of infrastructure investment should not be a reason for taking no action, infrastructure investments such as the Gander Airport and improvements to navigation such as the Tennessee –Tombigbee waterway make it clear that ill-conceived investments can be costly.

Improving the efficiency of current use lessens congestion and pushes back the time when a decision must be made on infrastructure expansion. As the Gander Airport example shows, delaying a decision for a few years could show that the investment should be designed differently or even is unnecessary. As soon as the Boeing 707 demonstrated that it was cheaper, more reliable, and much preferred by passengers, the airlines scrambled to buy jets and shed their propeller aircraft from their transatlantic traffic. Had they waited a few years before making their expansion decision, Gander planners would have seen that airport traffic was about to decline precipitously.

There are costs, however, to waiting for more information, just as there are costs to building a project on the basis of current information. If traffic continues to increase, waiting to build a project means that users will have to bear the costs of increased congestion for years. If traffic does not increase, as in the Gander example, construction costs will be mainly wasted.

Society needs a prudent rule for deciding when there is sufficient information to stop waiting and to start building. Such a rule involves tradeoffs between the social losses from building when it is not warranted, and delaying construction beyond the points where new capacity is needed. The decision also requires an estimate of the likelihood that the additional capacity is actually needed on the basis of currently available information. The latter is a technical decision; this committee can lend advice to the Corps on the estimation. The former is a value judgment that Congress—not the Corps—must make.

Extra throughput could be squeezed out of the current locks by improving congestion management. Although scheduling tows to arrive at the locks is difficult, valuable steps can be taken. The level of traffic on the Upper Mississippi River is not uniform over the navigation season. Smoothing traffic would significantly reduce congestion and delays. Nonstructural measures such as a scheduling system, congestion pricing, and tradable permits will also initiate smaller and far fewer environmental impacts than structural measures such as lock extensions. They are therefore more consistent with strategies for the sustainable development (promoting improved traffic flow and environmental restoration) of the Upper Mississippi River and tributary system.

When infrastructure investments are prudent, it is important to find the appropriate scale for the investments. Social resources are wasted by building a small structure that must be replaced in a few years, or by building a structure whose capacity is not needed. Thus, a careful analysis of the range of possible future demands and ways of accommodating them is needed.

The nation has undergone a profound shift in how environmental resources are regarded and valued. For example, the prospect of removing four large federal hydroelectric power dams on the Snake River in the Pacific Northwest, in order to restore salmon runs and the environment, has been discussed at the highest policymaking levels; several other U.S. dams have already been removed. These changes are occurring because environmental consequences have in some instances turned out to be worse than expected, and because society today places greater

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

value on not damaging ecological resources. The lesson is that infrastructure changes last for a long time, long enough for people's perceptions and preferences to change. In the committee's judgment, environmental concerns will continue to become more important in the future.

This chapter comments on these four analytical topics—making decisions under uncertainty, short-term nonstructural measures, longer-term measures, and environmental considerations and studies. It is provided to the Corps in the spirit of helping improve the draft navigation feasibility study.

CONTEMPORARY UNCERTAINTY ANALYSIS: THE STATE OF THE ART

Analysis Under Uncertainty

Complicating the Corps' analysis is the fact that infrastructure investments have long time horizons. About a decade would be required to extend the locks on the Upper Mississippi River–Illinois Waterway, after which they would be expected to last perhaps a century, with periodic maintenance. No one can know or predict with confidence the demand for water transport —or almost anything else—50 or more years in the future. In this section, we examine how to analyze uncertain future demands. Past Corps analyses did not recognize or treat uncertainty explicitly, even when dealing with shipments up to 50 years in the future. Failure to deal explicitly with uncertainty leads the unwary to have far too much confidence in the resulting forecasts and analysis, which can lead to bad public decisions concerning waterway investments.

Sensitivity Analysis

A common way to treat uncertainty is by conducting a “sensitivity analysis.” As an example, demand for grain shipments on the Upper Mississippi could be explored along the entire range of what is plausible in each year up to 2050. In particular, grain could be sold to local mills for processing or exported via a route other than the Upper Mississippi waterway. If so, barge shipments of grain on this waterway could be reduced to zero. At the other extreme, increases in world population combined with increased affluence and a growing demand for meat could increase world prices for grain to the point where much larger tonnages move on the rivers. Such a sensitivity analysis would find that the benefits of extending the locks would be small if grain shipments fell and would be substantial if traffic increased greatly. This method can be helpful if the full range of future values for the crucial variables is specified. It is easy to bias the analysis by artificially truncating the range by, for example, assuming that grain shipments could not fall. However, when the full range of future values is specified for all the important variables, the outcome of an uncertainty analysis is somewhat predictable. In general, one set of values produces demands too small to justify expansion while another set of values produces demands that make expansion a priority investment. Although the analysis can isolate the crucial variables and the crucial values of these variables, other methods must be used to assess the likelihood that demand will be high enough to justify lock extensions.

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Monte Carlo Analysis

A second method for treating uncertainty builds on sensitivity analysis by specifying a probability density function (distribution) for each uncertain variable. Unlike sensitivity analysis, which attempts to specify the plausible range, this second method—Monte Carlo simulation —attempts to place probability estimates on each value in the range. Once these distributions are specified, a Monte Carlo analysis is used to calculate the posterior distribution, given the structure and distribution of each variable. The Corps currently uses this method in its flood damage reduction studies (NRC, 2000).

If one has a good idea of the probability distributions for future values of each variable, and if those distributions can be specified with confidence, the resulting posterior distribution could be extremely informative. It would give the probability that the benefits would be high enough to justify extensions. Sensitivity analysis might conclude that some values of future demand are so low that lock extensions are not justified. In contrast, the Monte Carlo analysis might conclude, for example, that there is a 90 percent chance that lock extensions would be economically justified. This analysis would give the Corps and Congress information on the likelihood that lock extensions are justified—if there is general agreement on the probability distributions for the values of the crucial variables.

Unfortunately, analysts are hard-pressed to specify the range of uncertainty, much less the probability distribution. In these cases, analysts assume that the probability distribution is uniform or triangular or some other convenient, plausible, but unknown shape. In general, the posterior distribution and the probability that lock extensions are justified depend on both the range of uncertainty and the shape of the assumed distribution. Because the distribution is not known with certainty, the analysis should be conducted by examining all plausible distributions, a task rarely undertaken. The information that comes from a Monte Carlo (or any other type) analysis cannot be greater than the information and assumptions that went into the analysis. In other words, if one knows little or nothing, an analysis is not likely to provide a great deal of useful information. This obvious point, however, is often forgotten by analysts and by the sponsors of their analysis. If an analysis produces powerful results —despite the fact that little information was used as input to the analysis—these results should be viewed with some suspicion. In a Monte Carlo analysis, this might occur because the analysis unknowingly makes strong assumptions about the probability distributions.

Scenario Analysis

A third approach is the use of "conditional forecasts" or "scenario analysis." For example, one scenario would be expanding world demand for grain while another might be increasing competition from grain producers in other nations. Each of the scenarios is conditioned on (one assumes) a value for the crucial variables, and then explores the implications for lock extensions. In this sense, scenario analysis is a development of sensitivity analysis where particular values for each crucial variable are specified by sketching a particular future on the basis of some stated assumptions. The social gains and losses associated with lock extensions under each scenario

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

can be explored. If probability values can be placed on each scenario, the analysis becomes similar to the Monte Carlo analysis described above. If not, each scenario becomes a way of thinking about the future. Scenario analysis is more conceptual and qualitative than it is quantitative. It can produce insights but it rarely produces useful estimates.

Wait and See

A fourth approach is to recognize that the parameters really are unknown and that it is not necessary to make a final decision today. In this approach, the social costs and benefits of making a decision today are constantly evaluated against the social costs and benefits of waiting for more information.

But the future can hold discontinuous changes, such as catastrophic failures of infrastructure. When infrastructure will be subjected to the possibility of extreme events, engineers usually aim to account for such possibilities in project design. Dams, for example, may be constructed to withstand a “probable maximum flood.” However, decisions about appropriate design criteria, or when a project should be built, have little to do with “wait and see” analysis. Society should not allow crucial infrastructure capacity to lag behind demand and should not neglect inspection and repair on this infrastructure. Neither of these principles conflicts with a “wait and see” approach.

This approach becomes more attractive when steps can be taken to lower the cost of delaying the decision. For example, nonstructural improvements such as helper boats and improved scheduling, reduce traffic congestion at every level. Instituting such improvements means that the cost of delaying structural improvements is lowered, allowing more time to assess future demands for U.S. grain exports.

Finding Robust Strategies

Much of the future is unknown and unknowable. It would be wonderful if the Corps knew precisely how much barge traffic was going to use the UMR–IWW over the next 100 years. But they do not, and no amount of analysis can predict precisely the amount of future waterway traffic. Rather than designing the lock and dam system to minimize costs for a precise (and unknown) amount of future traffic, it would be prudent to seek a construction program that provides significant benefits for a range of future traffic levels, even if it is not optimal for any single amount of waterway traffic. Robust policies—those that produce favorable outcomes under the full range of plausible scenarios —should be sought.

This search begins with a scenario analysis to define the range of likely levels of future waterway traffic. The costs of a wide range of construction scenarios are then considered together with the congestion costs, and other social costs, for each traffic level for each construction program. The goal is to identify a construction program that produces reasonably high benefits or low costs over the range of plausible outcomes.

This approach can also be combined with a “wait and see” approach. For the range of

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

plausible scenarios, when will the true nature of future demand be revealed? In view of the costs of building before there is adequate information on the one hand, or the costs of waiting too long on the other hand, which alternative produces high benefits and low costs?

Comparing Approaches

Each of these approaches can be helpful, depending on what is known about the future. If a single variable is crucial, a sensitivity analysis can isolate the critical value that determines if expansion is worthwhile. If so, subsequent analysis can focus on the likelihood that this critical value is likely to be exceeded. If more than one variable is important and the analyst has a good idea of the distribution of each variable, a Monte Carlo forecasting approach can prove helpful. Just knowing that a small number of values of one or a few crucial variables are important can allow a scenario analysis that helps clear away confusion. Finally, finding a robust strategy that is beneficial across the range of uncertainties may require little information about the future. The best approach to each situation depends on the nature of the situation and what the analyst knows.

Generally speaking, some things are known about the future. A good deal is known about what the next few years will bring, but less is known as one looks further into the future. A major problem in uncertainty analysis is that analysts tend to be overconfident about their knowledge of the future. In many situations, actual events turn out to be outside the range of forecast futures.

All the options considered seriously by the Corps in formulating plans for navigation enhancement appear to involve large-scale lock and dam extensions and would require large public expenditures. A benefit-cost analysis to evaluate these expenditures requires information about waterway traffic over the next century; no one can predict future waterway traffic levels precisely. If demand for barge-transport capacity increases rapidly in the future, expansion will be in the public interest. But even current levels of waterway traffic generate enough congestion and delays so that some response is in the public interest: if the demand for barge transport continues as it has been for the past five years, no extensions will be needed.

Fortunately, several “management” improvements in tow operations could produce benefits at little cost: industry “self-help” efforts, incentive/disincentive fee structures, scheduling changes, speed limits, and installation of better barge coupling/decoupling equipment. Indeed, these nonstructural improvements not only can be made, but they should be required before determining whether and when construction is needed. The benefits and costs of nonstructural traffic management plans should first be carefully estimated before estimating the benefits and costs of lock extensions.

SHORT-TERM (NONSTRUCTURAL) MEASURES

Congestion Problems on the Upper Mississippi

On the Upper Mississippi River–Illinois Waterway, towboats arrive randomly at locks.

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

This is expressed through unpredictable and random, sometimes very intense, periods of congestion. There is no perceptible pattern to arrivals of towboats at locks throughout the day. This leads to queues at locks of unpredictable lengths. Thus, for example, a tow that arrived at Lock 25 on October 8, 1999, at 11:35 p.m. locked straight through. A tow arriving one day later waited in a queue of more than 22 hours. A tow arriving two days after that locked through with no waiting.

Motorists are familiar with predictable traffic congestion patterns and plan their routes and departure times accordingly. Although drivers cannot completely eliminate congestion problems, they can work around them by changing routes and departure times. By contrast, imagine that traffic congestion was frequent and random and that one could not predict when or where traffic congestion would occur and could not alter one's route. If one wished to be sure to arrive at the office by 8 a.m. every morning, it would be necessary to leave hours early each day in case of an hours-long traffic jam that could not be avoided. Although one would almost always arrive by 8 a.m., on many mornings one would arrive at the office by 7 a.m., and much of the time it would be closer to 6 a.m., depending on traffic levels. Clearly this would involve a substantial waste of time and resources.

This is the position that inland waterway operators find themselves. Through acting individually with no coordination, they have opted to treat lock congestion as an uncontrollable event. Congestion slows movement on the system and results in millions of dollars of wasted resources. This raises shipping costs on the UMR –IWW and raises the rates that operators charge for their services. Because farmers' incomes are determined by the difference between the world price of crops and the cost of delivering crops to market, lock congestion on the UMR–IWW reduces incomes of the agricultural community in the Upper Midwest. Service quality is diminished because operators cannot give a reliable delivery date and time, thus diverting traffic to other modes. The key to solving the congestion problem on the UMR–IWW is to reduce the randomness of arrivals at each lock. If arrivals at locks became more regular and constant throughout the shipping season, system capacity could be increased, traffic could move more quickly, and rates charged to agricultural interests could be reduced.

For example, large airports limit the number of flights that can be scheduled to take off or land during an hour. But in the 1980s, few airports used this type of scheduling, which resulted in long queues to take off and many flights in holding patterns waiting to land. When airlines and passengers realized the frustration, high costs, and danger of such an unscheduled system, they agreed to adopt scheduling. The current system naturally works best under perfect weather and other ideal conditions, but it also generally works well when affected by flights delayed by maintenance, bad weather that reduces airport capacity, and other unexpected difficulties. Many airlines would like to schedule an extra flight during the period of highest demand. They understand that this additional flight will face some delays because of congestion. However, this additional flight would also impose costs (in the form of additional congestion and delays) on other users. These types of external costs caused by congestion must be internalized to have the system operate efficiently. For example, if an airline had to compensate other users for the additional waiting time imposed on them by this additional flight during a high-use period, these costs would discourage the airline from adding the extra flight.

An airport takeoff or landing “slot” at a large airport during a busy time is a valuable

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

property. The federal government has essentially given airlines a property right to slots on the basis of when they first started scheduling flights in a given time window. Allowing airlines to buy and sell slots has enabled them to transfer slots to more valuable times and given them more flexibility in planning their operations.

Airlines, passengers, airports, and the federal government find it hard to believe that they tolerated the congestion that existed without the current scheduling system. The current system has enabled airlines to schedule their crews more predictably and efficiently, has reduced their costs, serves passengers better, has reduced frustrating congestion delays, and has increased safety. No one is interested in returning to the old system.

Precisely the same principles apply to tows seeking to transit a lock. Congestion delays are frustrating and expensive and increase shipping costs. While each tow operator would prefer to have first priority at each lock, they understand that this is not possible. Rather than waiting many hours in a queue, tow operators would prefer to know when locks will be free so that they can schedule their fleets to optimize equipment usage. For example, if an operator knew that there was a window of clear sailing next week from St. Louis northward to Keokuk for pickup, then back downstream after pickup, the towboat could be dispatched at that time, thereby allowing productive use of the equipment on the Lower Mississippi until the entry time into the lock system. This increase in the hours per year during which the towboat was productively deployed would allow towboat operators to amortize their equipment costs over a larger traffic base and could reduce commercial cargo shipping rateson the UMR–IWW. Unless the introduction of traffic management schemes changes the market structure of the barge industry—a change not likely to occur, based on both theoretical arguments and historical experiences—the reduced costs will be at least partially passed on to the consumer.

There is another advantage to greater predictability of arrival times. If operators knew that there were time frames within which clear sailing through all locks would be possible, priority services could be developed for waterway operation. The development of highly dependable, high-speed services for high-value freight was one of the unforeseen consequences of the deregulation of railroad and highway motor freight. Although waterway freight is typically low-value and is thus not as attractive to priority services, the possibility that there is a demand for highly predictable, higher-speed waterway services— a demand currently not satisfied because of unpredictable congestion delays at locks—cannot be ruled out.

Congestion has been analyzed in many systems, from highways to ports to supermarket checkout counters. Some standard ways of handling the externality are (1) charge a congestion toll equal to the social cost imposed by each additional user, (2) require that users make reservations for using the system to eliminate excess demand, and (3) ration the number of users by selling or giving away slots that enable individuals to access the system at a particular time (such as takeoff and landing slots at Washington's Reagan National airport).

Each of these approaches can increase the efficiency of the system and lower the social cost. Market-based systems promise greater efficiency than approaches such as “first-come, first-served.” Three issues are involved: (1) reducing waiting time at locks by better scheduling, (2) getting the tows with the greatest need for haste through first, and (3) minimizing the time each tow ties up the lock. A market system can provide incentives for improving all three aspects of efficiency. In contrast, a nonmarket system, such as first-come, first-served, can deal

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

with only the first issue. Allowing tows to trade slots can deal with the first two issues. To address the third issue, tows must face an incentive to minimize the time in the lock, such as through a charge for each minute spent in the lock.

One market-based system is a congestion toll, where each tow pays for the cost of the delays imposed on other tows. This system deals with all three issues, but it can be complicated and is likely to result in high tolls at congested locks. A second market-based system involves issuing a property right for the amount of lockage time used in a historical period (see Appendix C for a discussion of this system). The lockage time could be broken down into 5-minute increments. By trading the increments, a tow operator could assemble the total time required to get a tow through a lock at the time when the tow needs to get through. Because the time is in 5-minute increments, each tow operator is motivated to minimize the time spent in the lock.

Each of these market-based systems is more complicated than the current system. Towboat operators, lock masters, and other current users of the system may react negatively to proposals for such large changes in the system for managing lock operation. However, market-based systems have proven to be highly beneficial in many applications. And, in the committee's judgment, the Corps and tow operators could quickly learn to use these systems and benefit from their implementation.

The primary advantage of a system of tradable slots is that it would allow towboat operators to plan their season's sailing schedule in a way that leads naturally to a smoothing of arrivals. This would speed the flow of traffic, allowing more trips for each tow per year, and would save on towboat costs, resulting in lower rates to users. There are other advantages as well. The second most important advantage of the system is that it encourages efficient use of the locks themselves. There is a huge variation in the speed with which tows traverse locks.

For example, in October 1999 at Lock 25, the average time for a tow requiring a double lockage to transit a lock heading downstream was 78 minutes. 95 percent of the tows took 114 minutes or less, while 5 percent of the tows took 48 minutes or less. For tows heading upstream that required a double lockage, the average time was 92 minutes. 95 percent of the tows took 127 minutes or less, while 5 percent took 64 minutes or less. Thus, the fastest tows took half as long as the slowest tows. If all tows could be made as efficient as the fastest current 5 percent. many more tows could be served by each lock, essentially eliminating the current congestion.

A system of lockage permits would give an incentive for inefficient operators to increase the speed at which they cleared the locks. Those operators who were slow to clear locks would find that they needed to reserve more lockage time than the faster operators. The increased costs would come from an operator's bottom line because users would choose the shipping company with the cheaper rates. Operators would respond, for example, by increasing the training of deckhands to speed recoupling, by adopting new coupling techniques or, when appropriate, by paying another operator for assistance in clearing locks. It must be expected that operators would experiment with methods than are now not currently used to see if there would be savings in lockage times. Of course, if average lockage times are reduced, more tows per day can use each lock. It is impossible to predict the degree to which new technologies explored in the effort to economize on lockage time would expand the economic capacity of the current locks. If history is a guide, however, the saving would be substantial. Providing clear title with trading

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

privileges is one of a class of incentive-based policy instruments that could be used to induce more efficient use of navigation infrastructure. For example, incentive-based instruments have been adopted to address lead in gasoline, ozone-depleting chemicals, nitrogen oxide and sulfur emissions, new-vehicle fuel efficiency, urban land development, and retirement of older, heavily polluting vehicles. Emissions constraints and tradable permits have been proposed by the U.S. as a strategy for controlling global emissions of greenhouse gases. Incentive-based policy instruments include user fees, congestion pricing, pollution taxes, cap-and-trade policies, deposit-refund programs, and replacing subsidized prices for inputs and outputs with market-driven prices. Appendix C describes experiences in tradable permits in several recent applications outside of inland navigation.

These nonstructural options for improving waterway traffic management hold great promise for helping alleviate waterway congestion quickly, at relatively low costs, and in a manner that is more consistent with the long-term environmental sustainability of the Upper Mississippi River ecosystem than large-scale, structural changes. There are benefits and costs associated with each nonstructural option, the careful assessment of which was beyond this committee's scope and resources. As a first step toward improving UMR–IWW traffic management, a careful study of the benefits and costs of nonstructural options for improving waterway traffic management should be conducted.

Other Strategies for Alleviating Congestion

An alternative to offering transferable titles to lockage slots would be to allow an operator who was within three days of arriving at the first lock in a sequence to reserve lockage times at that lock and all succeeding locks to be traversed. This would increase the predictability of the lockages at succeeding locks. Any technique that increases the predictability of arrival times would help to eliminate wasteful development of queues at locks. We consider this system to be less desirable than a system of clear title to lockage times for several reasons:

  • It would be necessary to determine how long a lockage time should be. As noted above, some operators have longer times and some shorter. If lockage slots were determined by the longest time, the locks would stand idle following a fast tow, thus wasting resources. If the lockage time was determined by the fastest operator, it would be necessary to develop a penalty scheme to penalize slow performance.

  • Only limited smoothing of arrivals would take place. For example, an operator would have no assurance that he or she would not encounter a queue at the first lock. The operator would have no assurance that the desirable sequence of lockage slots was available upstream (for northbound movements, for example). Similarly, there would be less incentive for seasonal smoothing of demand.

  • An operator would have an incentive to over-reserve slots. For example, assume that a towboat operator planned to travel upstream from St. Louis to Quincy, Illinois. As the operator approached St. Louis, the operator would be tempted to reserve lockage slots all the way to Lock and Dam 1 at Minneapolis, in the hope of having a smooth trip all the way up the river. In order

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

to avoid such gaming of the system, a program of penalties for reserving but not using lockage slots would be needed. This complication would not occur if all operators were given clear title to lockage times at the beginning of the navigation season based on historical usage patterns.

  • Unlike the permitting scheme in which the bidding process for Corps-owned lockage slots provides a tangible and indisputable measure of willingness to pay (and thus a measure of the benefits of lock extensions), the value of lockage times would not be revealed by such a system. Users would thus be in the same uncomfortable position they are currently in—an inability to demonstrate, beyond a reasonable doubt, the advantages of lock extensions.

LONGER-TERM MEASURES

Traffic Demand Forecasts

Several Corps of Engineers papers have advocated the use of spatial equilibrium models to evaluate the economics of public investments. A 1998 report from the Corps' Economics Work Group, for example, states, “The consensus of this literature is that the economic impacts of transportation systems are best analyzed as components of larger spatial price economic models. Analyzing transportation systems or their individual components myopically can lead to erroneous conclusions regarding economic impacts and values of the transportation system and its components. Spatial price economic models may be characterized as models where consumer demands and producer's supplies of goods and services are identified by their location in spatially geographic regions called markets” (USACE, 1998). The committee agrees with this basic approach.

The above statements imply that an evaluation of investment in inland waterway navigation improvements should be based on disaggregated spatial equilibrium models where all relevant alternative supply and demand regions are identified and connected by product prices, alternative modes, and transport rates. Development of the disaggregated spatial equilibrium grain models involves five steps:

  1. The first step is to forecast world import demands and U.S. grain export demands, as nearly all grain shipments on the Upper Mississippi River are destined for export markets. This requires forecasting both the demand for imported grains and the supply of exported grains. The former depends on population growth, per capita income, and the demand for meat. The latter depends on grain production in Argentina, Brazil, the U.S., and other grain-producing countries.

  2. The second step is to forecast the amount of grain that each farmer will want to send to each market, e.g., local animal feed, processing, and exports. The expected grain price in each market, together with the shipping cost, will determine in which market and by which transportation mode the farmer can expect to get the greatest revenue.

  3. The third step is to aggregate the net revenue maximizing decisions of individual farmers in order to calculate the market equilibrium for both the uses of the grain and the shipping modes. This is because the price for grain in each market will fall as farmers desire to ship more grain to this market, and the price of transportation will rise as congestion increases because of increased shipments.

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
  1. The fourth step is to forecast the performance of each mode and route that is affected by the traffic volume carried on that link. The objective is to determine how the level of traffic affects congestion levels experienced by each link, and thus how rates charged by transport operators on those links change with traffic levels.

  2. The final step is to estimate the net benefits of transportation infrastructure investments.

These steps require the development of forecasting and optimization models.

Forecasting World Import and U.S. Export Demands

There are two basic types of forecasting models: structural models and nonstructural, or time series, models.

Structural models attempt to explain the behavior of dependent variables as a function of one or more causal independent variables or equations. The two basic large-scale agriculturalsector structural models are the U.S. Department of Agriculture (USDA) agricultural-sector model and the Food and Agricultural Policy Research Institute (FAPRI) model. Both models were designed to make projections rather than forecasts; that is, they were designed to evaluate alternative agricultural policies. Their focus is to measure the difference between alternative policy options rather than to forecast absolute levels of specific outcomes. To be used in forecasting, these models must be reformulated by replacing the largely linear regression equations with more modern forecasting equations like exponential smoothing and moving average equations. These latter equations place greater weight on recent data. In contrast, regression equations weigh 10- to 20-year-old data as equal in value to recent data.

Nonstructural, or time series, models do not attempt to use theoretical relationships such as supply and demand. Instead, they treat the variable of interest as an element to be forecast from past data. Several types of time series models are available. These include Box-Jenkins, ARIMA (autoregressive integrated moving average), and ECM (error correction models). These models typically consist of a time trend, a seasonal factor, a cyclical element, and an error term. Both types of models are likely to be superior to a linear regression time trend. It is difficult to incorporate nonquantifiable variables into all quantitative models. Therefore, judgment must be used to adjust these models to recognize nonquantifiable information. Both approaches can be useful in forecasting exports a few years into the future. Time series models are most accurate in extremely short-term forecasting, while structural models are likely to provide better forecasts of conditions beyond about two years into the future. Unfortunately, forecasting 20–100 years into the future is beyond the capability of these models. Instead, the forecasts depend on population growth, increases in per capita income, the demand for meat, and advances in agricultural productivity, all of which are inherently difficult to predict.

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Forecasting Model Accuracy

There are several methods of measuring model accuracy. One of the simplest ways is to compare model forecasts with actual values. Models can be used to forecast recent year outcomes, and these forecasts are then compared with actual outcomes. The actual values can be regressed on the forecast values and R2 (R2 measures the proportion of the variation in the dependent variable, e.g., shipments, accounted for by the regression) is used to measure accuracy. The higher the value of R2, the more accurate the model.

Dealing with Uncertainty

There are large uncertainties in making long-range export forecasts. One method of obtaining the “best” forecast is to use several forecasting models. Then, after comparing forecasted values with actual values, select the 2 or 3 highest R2 “most believable” models and calculate a “combined” forecast, formed as a weighted average of the 2 or 3 selected forecasts. The weight can be obtained from the regression of the actual values on the forecasted values. After selecting the supply origins—most likely individual countries or groups of counties—the quantities of grain available for shipment can be estimated by subtracting the amount of grain fed to livestock from the production in each supply region. Several types of models can be used to forecast the quantity of grain to move to each market by mode and time period. Among the simpler methods are the linear programming (LP) and quadratic programming models.

Data Requirements

Sweeney and Marmorstein (1998) suggest that the reason the ESSENCE model failed to include grain origins and quantities, alternative markets, product prices, and transportation rates from all origins to all destinations was because of a lack of adequate data. This section outlines the data necessary to run a disaggregated spatial equilibrium model for grain.

Origin Data The farmer makes the basic decisions that determine grain flows. A farmer's decisions on where to deliver grain from the farm initiate the grain distribution process. However, modeling every farm within the relevant river basin study area would be very expensive. A more practical approach is to define the origins as one point in each county or group of counties. The quantities of grain available from each defined origin can then be estimated.

Transportation Rates Grain sales and shipment decisions are made on the basis of commodity prices and current transportation rates, not on long-run railroad, barge, and truck marginal costs. Characterizing current farmer decisions and shipments to market requires knowing the prices that farmers have to select from. Although long-run marginal costs may be a rough approximation for truck rates, and a less good approximation for barge and rail rates, actual rates are needed to explain current farmer decisions and to estimate price sensitivities.

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

Three caveats are in order when collecting data. First, models should allow for barge and truck shipping rate seasonality by including several—preferably 12—time periods. Barge and truck shipping rates are constantly changing. Thus, attempts to remove “abnormality” from the rates by eliminating variability—are improper. In some supply regions, rail shipping rates are frequently lower than barge shipping rates. Secondly, barge, truck, and rail rate data should be for the same time periods during which the quantity data are collected. Finally, rates should be for specific days, rather than averaged over long time periods.

Quantities Shipped from Each Origin. Quantities available to be shipped from each origin—a county or group of counties—should be calculated by subtracting the amount of grain fed to livestock or otherwise used within the origin area.

Destination Data

Alternative Markets. A spatial equilibrium model must identify and incorporate all relevant markets. These markets can be identified in several ways. The best method is to conduct grain flow surveys in each state. These survey data are useful not only for identifying relevant markets, but also for calibrating modal grain flows and modal shares. In addition to asking barge terminal operators, another option to identify relevant markets that should be included in the model is to ask the grain shippers who make the sales decisions.

Market Prices. Market prices are just as important as transportation rates in determining grain flows among the competing markets. These bid prices, for days that must exactly match the dates for the rates and quantities, can be obtained from grain buyers.

Market Constraints. Most domestic grain markets face quantity constraints for given time periods. These constraints are based on daily processing and storage capacities. Animal feeding markets face both minimum and maximum quantity constraints. The maximum constraint is based on the number of animals fed, and the minimum constraint is based on the fact that the market must have grain supplies each day or the animals will starve. Data on processing market constraints can be obtained from newspaper articles and by asking industry people about processing capacities.

There should be no constraint on export markets other than the export projections. However, there may be environmental or governmental policy constraints on total imports into individual countries. These constraints must be built into the export projections.

Developing a Disaggregated Spatial Equilibrium Model for Grain Transport

A spatial equilibrium model seeks to find the most profitable destination and transport mode for grain from each area. These models recognize explicitly that grain is grown in tens of thousands of locations and that there are hundreds of markets for the grain in the form of individual feedlots, processors, and export locations. Such models are most accurate when operated

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

on a detailed spatial scale, such as a few farms or a county. These models summarize the historical relationships between the quantity of grain shipped by barge and barge shipping rates, the rates for other modes, and the prices offered for grain in the various domestic and export markets. A model developed on historical data can be used to forecast future barge shipments on the basis of assumptions about the future rates for each transport mode, the amount of grain grown in each area, and the quantity of grain used as feed, processed, or exported.

All participants in the grain distribution system are assumed to seek maximum profits. Therefore, the spatial equilibrium model—a programming, network or other disaggregated optimization algorithm —seeks the shipping routes and modes that maximize the total profits of farmers and transporters. The following steps should be used in developing a disaggregated spatial equilibrium model for grain:

Step 1: Identify the relevant geographic grain production areas to be included in the model. This should include all areas that ship grain on the Upper Mississippi River at least once a decade. These geographic areas should be subdivided into the largest number of (grain) origin areas that can be handled easily by the program (down to individual farms, if possible). The areas should be no larger than a county. One location within each area should be identified as the representative shipping origin for all grain to be shipped out of the area. The quantity of grain to be shipped out of the area should be defined as production minus the amount consumed within the area as animal feed or processed in mills; alternatively, within-area uses could be included as competing demands for the grain).

Step 2: Collect delivered grain prices for all destination markets to which this grain might realistically be shipped. These prices should be collected for one selected day for each month in the year(s) to be included in the analysis.

Step 3: Collect transportation rates for each relevant mode of transport from each shipping origin to each destination market for the same days on which delivered prices were collected. These transportation rates and grain prices would be used to calibrate the model to replicate actual grain flows for the year(s) from which prices were selected. In some studies, long run marginal costs (LRMC) for barges, railroads, and trucks have been used as surrogates for actual shipping rates. Unfortunately, actual shipping rates can be much more or less than LRMC at particular times because of circumstances such as high or low demand, heavy congestion, or weather-related difficulties. A model based on LRMC would fail to explain the drop in barge traffic when shipping rates were much greater than LRMC and vice versa.

The committee strongly recommends that the Corps obtain and use actual shipping rates—not published tariffs—and corresponding quantities shipped for each model.

Step 4: Use the programming, network or other optimization model to find the grain flows and transport modes that maximize the system profits—profits to farmers, shippers, and carriers—in shipping to domestic users and U.S. export ports. Market prices and transport rates should be used to calibrate the model based on past grain flows. To validate model results, the grain flow forecasts from the model should be compared with historical data (ideally, data not used to calibrate the model). The model should then be adjusted to correct for major discrepancies between the model and survey results. Changes in barge and other transport rates and in grain prices for various uses and destination can be used to estimate how sensitive is the demand for barge transport to various prices.

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

Step 5: Forecasting future benefits. While forecasting traffic flows depends on forecasting freight charges, the estimation of future benefits depends on long-run marginal social costs (LRMSC), which are defined as private (costs to the carrier) plus public costs. In forecasting future benefits, long-run marginal social costs should be forecast for each mode. The price for each mode will be long-run marginal social costs, plus or minus a factor that depends on the level of demand. The relationship between long-run marginal social costs and freight tariffs can be estimated with data from the past decade or so. The same relationship between long-run marginal social costs and shipping rates can be assumed to hold for the future. Public costs for barges include annual government expenditures on maintenance and operations, and Corps of Engineers expenditures for administration and research. Public expenditures for railroads should include annual government expenditures for research and labor retirement and federal, state and local subsidies for short line railroads. Public costs for trucks include damage to highways in excess of fuel and use tax collections, and research expenditures. These and other such costs should be included in long-run marginal social/transport costs.

Step 6: The long run marginal private and public modal costs should be adjusted to reflect the wide swings in transportation rates within and between years. Rate data should be collected in past years of high and low transport demand. These rate data from the past give perhaps the best indication of whether actual shipping rates will lie above or below long-run marginal social costs in the future.

Step 7: The effects of nonstructural improvements on the cost of barge freight should be incorporated into the long run marginal private and social barge costs. Then the spatial equilibrium model should be run to estimate a base solution with which to compare the cost savings of structural waterway investments.

Step 8: Using the traditional Corps method of comparing the benefits of waterway grain transportation, run the base solution model and the after investment model to estimate the benefits of waterway investments under alternative scenarios of high and low demand and high and low transport rates—which are calculated from the long run private and public cost estimates.

A spatial equilibrium model that relates actual traffic flows on each mode to the relevant shipping prices and local demands would be a large contribution to the Corps, farmers, and exporters. Not only would it specify the relationships between relative shipping rates for each mode and its level of traffic, it would specify traffic flows, document market volume, and provide myriad useful results such as: quantities shipped from each origin to each market, by mode(s); changes in shipments by origins; changes in receipts by destination; and shipments by mode(s) of transport. The price sensitivity of the quantity of grain shipped by barge could also be estimated from these model results.

ENVIRONMENTAL ANALYSIS

Adaptive Management

The emerging concept and practices of adaptive management are gaining currency within natural resource management programs and with many scientists and scholars. The principles of

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

adaptive management grew from perceived shortcomings of traditional resource management practices, especially in large, complex ecosystems characterized by a high degree of uncertainty and occasional “surprises” such as floods or fires.

Many of the adaptive management concepts being applied and refined today can be traced back to research initiatives in the 1970s at the International Institute for Systems Analysis (IIASA) in Laxenburg, Austria and at the Resource Ecology Center at the University of British Columbia. A widely cited document that resulted from the IIASA studies is Adaptive Environmental Assessment and Management (Holling, 1978). That volume identifies several key principles of adaptive management and emphasizes the interplay of ecological and social systems in enhanced understanding and improved ecosystem management. It also emphasizes the importance of the flexibility of resource management policies, and it promotes an iterative process in which policies are continually updated as knowledge of environmental behavior is gained. As Holling states, “the process of adaptive management and policy design integrates environmental with economic and social understanding at the very beginning of the design process, in a sequence of steps during the design phase and after implementation ” (Holling, 1978). Some of the components of an adaptive management approach include: (1) a shift from “trial-and-error” management to formal experimentation, recognizing that all management policies are experimental in nature, (2) careful monitoring of ecological and social effects and responses to management actions, (3) an integrated scientific approach to resources management that encompasses ecological, social, and economic considerations, (4) the investigation of uncertainties through formal experimental design and hypothesis testing, and (5) a commitment to ongoing adjustments in resource management based on the results of prior experiments (see also NRC. 1999b).

Given this topic's complexity and richness, specific definitions and practices of adaptive management continue to be refined. Another relevant definition of adaptive management is “a systematic process for continually improving management practices and policies by learning from the outcomes of operational programs” (Nyberg, 1998, p. 2). The concept continues to attract great interest, as evidenced by the adaptive management approach to resource management being applied by the U.S. government in the Pacific Northwest (Volkman and McConnaha, 1993), in the Florida Everglades, and in the Colorado River below Glen Canyon Dam (NRC, 1999b).

Adaptive Management on the UMR–IWW System

The UMR–IWW presents a unique challenge and an opportunity in which to apply adaptive management principles described by, among others, Holling (1978), Walters (1986), Lee (1993), and Gunderson et al. (1995). There have been no formal efforts at merging adaptive management principles with navigation system operations on the Upper Mississippi or on any other of the nation's navigation systems. Nonetheless, several elements of the UMR–IWW and the navigation feasibility study suggest that an adaptive management framework could help provide unity and cohesion to the feasibility study.

One principle of adaptive management is to avoid premature foreclosure of management

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

options in order to avoid the trap of irreversibility: “We cannot always require a complete return to starting conditions or complete freedom to reach any other conceivable situation. But we can try to keep from getting locked into any one situation. No guarantees exist, but to ask honestly what options are being foreclosed reorients the planning and development process and makes dead ends less likely” (Holling, 1978).

In the Corps' feasibility study, the extension of locks is contemplated. Conducting an assessment of the benefits and costs of nonstructural options before considering lock extensions would provide the Corps additional time to study and learn more about the environmental effects of society 's activities on the Mississippi River and throughout the river system.

An adaptive management approach would place environmental and social considerations on par with economic considerations. As Holling stated, “Environmental dimensions should be introduced at the very beginning of the policy process design, and should be integrated as equal partners with economic and social considerations” (Holling, 1978). Given the links between ecology, society, and economics in the Upper Mississippi region, an adaptive management strategy would aim to balance these considerations at the outset of an investigation regarding an expanded navigation system. Encompassing these different factors within a multidisciplinary decision making paradigm is also consistent with the notion of sustainable development. In a more recent volume on adaptive management (Gunderson et al., 1995), Holling states: “Sustainable development is neither an ecological problem, a social problem, nor an economic problem. It is an integrated combination of all three” (Holling, 1995).

Adaptive management entails careful environmental monitoring and assessment in order to determine how management policies affect environmental systems. In the UMR–IWW system, further analysis of how the navigation system has affected river ecology and would be consistent with an adaptive management strategy. Within the EMP for the Upper Mississippi River. the Corps and its cooperating federal agencies continue to enhance knowledge of how the navigation system has affected river ecology.

Adaptive management does not avert the challenge of forging a workable vision for the desired state of the ecosystem and the balance between environmental quality and human use of resources. Adaptive management ideally begins with a set of management objectives and then experiments to observe how those objectives may be better and worse served by various practices. The absence of well-defined management objectives is a common criticism of adaptive management efforts. This criticism would also haunt adaptive management efforts on the UMR–IWW if clear environmental quality and navigation use objectives are not established. Effective adaptive management for the UMR–IWW system would require a sustained effort to define management objectives (including input from the public and formally incorporating public opinion into management objectives) and to identify which ecosystem indicators and which human uses of the waterways will be included in adaptive management experiments. Through numerous public meetings in which it discussed the feasibility study, the Corps has laid the groundwork for meaningful public involvement.

Adaptive management is a long-term process and can be costly—both in terms of monitoring and research and in terms of the opportunity costs of deferring management changes until such changes are evaluated by adaptive management experiments. If adaptive management is to be effectively implemented in the Upper Mississippi River, it will require a sustained level of long-term funding for ecosystem monitoring and evaluation.

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

The resilience of the politics and institutions governing the resources at stake is another important consideration. Are these governing structures likely to allow changes in resource management in response to a new understanding of science-policy relations based on results from ecosystem monitoring and evaluation?

Independent review of scientific results and their inferences for policy innovation is essential to the integrity of ongoing adaptive management programs. Without such independent review, adaptive management processes can be circumscribed by stakeholder concerns, and the most essential scientific questions may be set aside because of political pressures. Independent review of scientific findings is also valuable when policy recommendations for promoting adaptive management are unpopular with some stakeholders. Several review “teams” are already in place on the UMR–IWW feasibility study, although the committee recommends that external, expert review—independent of potential conflicts of interest—be conducted as part of the UMR–IWW feasibility study.

Given the potential for an adaptive management strategy to help manage the multiple resources in the Upper Mississippi River Basin, the Corps should pursue the navigation feasibility study within the principles of an adaptive management framework. The Corps should seek the authority and funding from Congress necessary to conduct the study within an adaptive management framework.

The Corps' Adaptive Mitigation Strategy

Absent a fully implemented adaptive management approach, the Corps describes its approach to mitigating the environmental effects of incremental increases in waterway traffic as “adaptive mitigation. ” Although the introduction to the Corps' adaptive mitigation strategy demonstrates confusion over the respective definitions of adaptive mitigation and adaptive management, with these terms being used interchangeably (USACE, 2000b; preliminary draft EIS, pp 6–5), the Corps clearly recognizes the value of adaptive approaches. However, a strategy that merely seeks to mitigate future negative environmental effects—without some type of environmental assessments and feedback during the expansion of the navigation system —does not reflect the principles of adaptive management as defined by Holling (1978), Walters (1986), Lee (1993), Gunderson et al. (1995) and others. For example, the feasibility study treats environmental issues and resources as planning constraints that require mitigation activities, rather than as valued resources that are to be protected or enhanced. The practical steps outlined in the Corps' adaptive mitigation strategy are generally reasonable; however, the committee rejects the notion that environmental resources should be treated only as constraints in the proposed expansion of the navigation system.

Improving the UMR—IWW Environmental Analyses

A first step toward improving environmental studies in the UMR–IWW feasibility study would be to take actions to address criticisms from local, state, and federal environmental interests. When considering the committee's recommendations for improving environmental analysis, it

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

should be recognized that the feasibility study was bound by earlier decisions that limited analysis to incremental effects and focused analyses to site-related project impacts. Furthermore, the feasibility study is being conducted in the context of a long history of environmental investigations that have included ecological research on the UMR –IWW system, assessments of impacts from the second lock at Lock and Dam 26, and Environmental Management Program data gathering efforts. Finally, time is also important when considering both the sequence of project activities and the overall project planning period of 50 years. Sequence is important because protection of both ecological resources and navigation capability in the Upper Mississippi River basin depends on a knowledge base for ecological impact analysis and for construction. Clearly, if more is known about ecology, that understanding can be used with knowledge of actual construction practices and construction activity timing to avoid or mitigate impact.

Important long-term planning considerations also include not only how impacts may be distributed in time, but also how anticipated changes in towboat numbers relate to possible time-related external changes. For example, a changing climate may alter stream flow, annually and seasonally. Clearly, changes in water availability will affect both ecology and navigation. Another example is land-use and population changes. Land-use changes can affect water quality and management requirements for ecosystem protection in the Upper Mississippi River basin. Changes in population may alter use intensity and traffic patterns. A comprehensive water resources planning analysis would consider these and other related issues in an environmental impact assessment.

Several recommendations aimed toward improving environmental analysis per se are provided in this section and discuss more thorough inclusion of those analyses in navigation project planning. The main analyses the committee focused on were a systemwide analysis of the navigation system's environmental effects, analysis of the cumulative effects of the navigation system on river ecology, and site-specific analyses.

Systemwide Analysis of Environmental Effects

There has been a long history of ecological research and analysis for the Upper Mississippi River basin, and the Environmental Management Program has established the Long Term Resource Monitoring Program (LTRMP), yet gaps in knowledge of the large, complex Upper Mississippi River ecosystem still exist. Some of those data gaps were expected to be filled by the “second lock” studies, which recognized that a detailed study is needed to characterize existing ecological status and to support impact predictions. For example, there is not a thorough understanding of how fish populations are influenced by navigation, land-use changes, and population growth in the Upper Mississippi's sub-basins. With such data gaps, full assessment of the incremental effects of proposed lock extensions on the environment is not possible. If population size and dynamics of fishes cannot be accurately assessed, how can risk predictions based on the expected mortality of fish caused by towboat passage be validated? It is simply not possible to calculate a percent loss in population if accurate estimates of total population size are not available. Collecting the data necessary for such a comprehensive analysis is beyond the scope of the Corps ' 1997 Project Study Plan, but these data are needed to support

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

the impact assessments required of the feasibility study. In the 1986 Water Resources Development Act, Congress established and provided initial funding for the Environmental Management Program. The Environmental Management Program has been an excellent investment of resources into the investigations of the Upper Mississippi River's ecosystems and the continuing effects of navigation improvements and other human activities. Congress should provide continued funding to the EMP, with specific and continuing appropriations for ecological studies (including cumulative effects analysis), and navigation effects studies, in the UMR–IWW feasibility study.

Cumulative Effects of the Navigation System on River Ecology

Impacts to river ecology are seldom attributed to a single cause; rather, many small effects from multiple causes typically accumulate to produce a damaging impact. A cumulative effects analysis accounts for these multiple causes. In the Upper Mississippi River Basin, the navigation system, and the operation and maintenance of its locks and dams, are affecting river ecology in multiple and complex ways. There is not a thorough understanding of how current operations (e.g., intra- and inter-annual changes in navigation pool elevations) and maintenance (e.g., dredging, construction of wing dams) of the navigation system, as well as other factors such as changes in land use and water quality, are affecting river ecology. This understanding is essential to an assessment of how future changes in the navigation system might affect the environment.

To address this need for enhanced understanding of these cumulative effects throughout the UMR–IWW system, a good starting point would be a detailed assessment of how current operations and maintenance activities, when combined with environmental changes, are affecting the environment. As part of the feasibility study, an analysis of the cumulative effects of the existing navigation system should be conducted. This analysis should account for project maintenance requirements based on construction methods, and—considering the duration of the project elements—it should assess maintenance and operational requirements in light of possible changes in external factors such as climate, land-use changes, or changes in population. This cumulative effects analysis should account for all major factors that can create significant environmental impacts in the UMR–IWW system. Examples of these major factors in the Upper Mississippi River Basin include spills of chemicals and fertilizers shipped on the river, changes in navigation pool elevations, and physical changes to the river channel.

The committee also recognizes that a historical context may be important for some navigation system planning investigations. The USGS “Status and Trends” report (USGS, 1999), is a good example of effective use of historical data. Ongoing studies of the UMR–IWW ecosystem continue to enhance the understanding of how river ecology is changing and responding to impacts over time.

Within this cumulative effects analysis, two additional sets of more focused analysis should be conducted: recent navigation system improvements and increased towboat traffic.

Recent Navigation System Improvements. In addition to the lack of understanding of systemwide, cumulative environmental effects, there have also been concerns that the impacts of recent navigation system improvements on the UMR–IWW are not fully understood and that they have not been adequately assessed. In fact, some groups (e.g., NECC, USFWS) view continuing

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

degradation, habitat losses, and the decline of select native species in the Upper Mississippi River basin as evidence of accumulating impacts (USGS, 1999). They argue that any assessment of impacts in the feasibility study, without a better understanding of both present conditions and trends, will be unfounded. The committee agrees with this view. Cumulative effects analysis on the UMR–IWW should include a systemwide assessment that also considers the continuing effects of recent navigation system improvements (e.g., siltation of backwaters due to navigation pool formation).

Increased Towboat Traffic. Improvements to navigation systems and increases in towboat traffic will generally produce site-specific and passage-specific effects. These effects are analyzed in the Corps' draft feasibility study and are considered in the Corps' draft EIS. An inadequacy of the feasibility study is the lack of analyses that consider cumulative effects of increased towboat traffic. There is a need to establish how elements of the proposed navigation system improvements—when coupled with increases in waterway traffic—might combine to produce a total environmental impact greater than the sum of its individual parts. For example, the Corps' 1997 Project Study Plan identifies multiple locations where construction activity is anticipated in the channel. A reasonable question is, “Might multiple improvements, constructed in sequence over the life of the project, have a total impact that single or isolated projects do not?”

In the feasibility study, although a range of options and scenarios has been evaluated, consideration of the cumulative effects of increased waterway traffic is inadequate. Further, cumulative effects may have sources that are directly related to the presence of towboats. For example, one might ask, “What will be the cumulative effect of a spill of oil, fertilizer, or toxic chemicals on critical periods in an organism's life history, given increased traffic movement, altered pool management, or the timing of the spill?” Questions such as this have not yet been addressed in the feasibility study, yet are important to a comprehensive understanding of how multiple changes in the river system might affect the environment.

Site-Specific Impact Assessment

The impact assessments performed as a part of the feasibility study draft EIS are of necessity general at this stage of project planning. This generality arises from the fact that several alternatives are still being considered. Considering these many alternatives, any impact analysis must be somewhat general. Unfortunately, this generalized impact assessment approach will of ten be inadequate when site-specific project activity commences. For example, detailed placement of structures, decisions on construction methods, and timing of construction activities are all uncertain at this time. Each of these factors can play an important role in actual impact. It is important to regularly review and update site-specific impact assessments as the project progresses. It is also important to recognize that each impact assessment activity can contribute valuable data to baseline data resources. It is important that impact analysis activities be coordinated with research centers so that maximum utility can be obtained from any assessment investments in the UMR–IWW system.

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

Detailed, site-specific impact assessments of construction activities on the UMR–IWW should be an ongoing activity in the feasibility study. It is particularly important that as experience is gained from early stages of the project, the experience contribute to better impact assessments. Furthermore, the Corps should develop a mechanism that will assure coordination with research centers in the Upper Mississippi region so that assessments will not only serve to meet impact analysis requirements, but will also contribute to improved understanding of the Upper Mississippi River ecosystem. Finally, these site-specific impact assessments should have a coordinated regulatory review. This site-specific impact assessment need not be a barrier to project progress; rather, assessments should be considered an essential element of site-specific design activities that will assure that construction effects are minimized and impacts truly avoided.

Congress should supply the resources needed to support the committee 's recommendations for additional environmental investigations on the UMR–IWW. The Corps should not be required conduct these analyses alone, but rather the analyses should be conducted with the cooperation of federal and state partners who have contributed to the feasibility study. The cumulative effects and other analysis should build upon the numerous, significant research efforts on Mississippi River navigation and ecology, including the Corps' draft Environmental Impact Statement.

Toward Better Integration: Interagency Coordination

Since the 1986 Water Resources Development Act, considerable progress has been made in improving ecological understanding of Upper Mississippi River ecology. The final documentation produced by the Corps of Engineers should integrate findings from the Environmental Management Program, not only in a draft environmental impact statement, but with all elements of the feasibility study (economic, engineering, and ecological). To date, the EMP documentation stands alone, as do many of the studies performed for, and by, the Corps as part of the feasibility study. But more than an assembly of these studies is needed. Integration is not simply the collection of studies into a single report; rather, integration involves a careful analysis of findings that connects and harmonizes individual elements of the feasibility study. The current challenge to the Corps is to continue what has been started and effectively integrate all project elements within the final feasibility and environmental impact analyses. Considering the schedule and progress of the feasibility study, there has been limited time devoted to achieving final integration. The Corps should take advantage of any opportunity to secure the time needed to complete an integration of feasibility study elements with EMP studies, and with environmental analyses available from Upper Mississippi states.

Toward Better Integration: Interdisciplinary Considerations As indicated earlier, the Corps has embarked on a landmark assessment in the feasibility study for the UMR–IWW system. Detailed economic, engineering, and ecological analyses, along with modern planning and assessment approaches, have been the foundation for this project. Although

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

progress has been made in analytical areas, there has been little time to use the findings from one set of analyses with other analyses (economic, ecological, or engineering). It is imperative that findings from each area of analysis are effectively integrated with other analyses in final Corps feasibility and impact assessments. For example, if economic modeling suggests a change in seasonal demand that will alter the number of towboat passages along a reach of river, then specific environmental impacts associated with those changes should be assessed. If shipping schedules are altered because of economic incentives, then the altered schedules should be assessed in terms of seasonal ecosystem dynamics. If engineering plans call for winter construction, seasonal impacts of that construction should be an element of integration. These are just three of many possible examples that could be cited as needed integrative analysis in the feasibility study. The Corps should aim toward a more comprehensive and integrated assessment of the effects of changes in barge traffic on the UMR–IWW. This assessment should be consistent with National Economic Development (NED) accounting procedures and environmental quality accounting procedures.

Environmental Costs and Valuation of Ecosystem Services

An acceptable approach for addressing environmental costs would be to detail the full range of environmental impacts associated with each alternative, and to estimate economic costs associated with these impacts. Only after this analysis has been performed should the potential for mitigation and the costs and effectiveness of alternative mitigation strategies be considered. Explicit consideration of environmental costs prior to selecting mitigation strategies allows for trade-offs between unmitigated environmental costs, and the costs of mitigation to alleviate those costs, to be examined. This approach will assist in setting priorities for mitigation and in cost-effective use of mitigation funds.

Many traditional measures of productivity and the values of goods and services, such as Gross Domestic Product (GDP), do not account for the values of nonmarket goods and services. For example, the value of a forested bottomland sold as timber (market value) would be included in GDP calculations, but not that bottomland's value in attenuating floods flows or in filtering pollutants (nonmarket values). In response to this shortcoming of many traditional measures and methods, an abundant literature on the valuation of nonmarket values for environmental goods and services has blossomed.

Several methods for valuing these “ecosystem services” have been developed and include the travel cost method, the contingent valuation method, and hedonic valuation. These economic methods are well developed and are used by many public agencies. While development of these methods dates back decades (Ciriacy-Wantrup, 1947; Davis, 1964), the Exxon Valdez incident in the Gulf of Alaska in 1989 generated a great deal of interest and activities in nonmarket valuation of environmental goods and services, specifically in contingent valuation methods (Arrow et al., 1993; Carson et al., 1994). While economists and others continue to debate the utility of contingent valuation methods (Hausman, 1993), contingent valuation is a widely used tool for valuing environmental goods and services. These methods aim to understand how much money individuals would be willing to pay for successive additional quantities of a nonmarket good. Although

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×

no method is foolproof, this value can be captured through questionnaires, surveys, and votes (see also Schelling, 1968). These methods today are used around the world (Hanemann, 1994) and are continually being refined. Of recent studies that have attracted attention, a group effort in the late 1990s that attempted to value all of the planet 's ecosystem services and natural capital stood out (Costanza et al., 1997). This 1997 paper was quick to point out that, due to the study 's complexity and ambition, that it was merely a first approximation of these values. In the field of water resources planning, there is a large and growing literature on valuation of the services of riverine and aquatic ecosystems (e.g., Bishop et al., 1987; Douglas and Taylor, 1998; Loomis, 1998).

These valuation methods, while essential in a complete examination of costs and benefits of project alternatives, cannot be usefully applied in the absence of well-defined descriptions of the range of environmental impacts that are associated with project alternatives. Moreover, these valuation methods will not be capable of quantifying (in dollars) all the environmental costs associated with alternatives. They are best suited for quantifying costs associated with environmental impacts that relate to human use of affected resources (such as boating and sport fishing) and impacts that can adequately be described to the public in the context of survey instruments of the type used in contingent valuation.

Recreation. A 1995 study sponsored by the Corps documents over 12 million visitor days annually, with associated annual economic impacts of $1.2 billion in recreation activities directly linked to uses of the UMR–IWW system (Carlson et al., 1995). Although they are not addressed in the feasibility study, recreation user benefits (consumer surplus to recreational users) are generated by these user-days (as are the economic impacts on businesses cited above). Direct benefits to recreational users are part of NED and should be considered in project evaluation.

Prior research suggests that river recreation is negatively affected by increased commercial barge traffic (Becker, 1981; Graman et al., 1984). In particular, recreational boaters respond to increased traffic by foregoing recreational boating and by relocating their activities to other sites that are not affected by higher traffic (with higher travel costs incurred and lower consumer surplus generated). This type of effect on recreation is consistent with studies of other forms of recreation that indicate a negative relationship between increased congestion and user benefits. Analogies can be found in the literature which documents negative impacts of increasing numbers of users at parks, fishing sites, reservoirs and hiking trails on recreation user benefits (Loomis and Walsh, 1997; pp. 104–109). The recreation economics literature clearly demonstrates that benefits per user-day fall notably as the number of encounters with other parties rises. This phenomenon should be incorporated in the analysis of project alternatives, which should explicitly model and quantify the relationships between recreation use benefits for boating and fishing and changes in commercial barge traffic levels.

River-related recreation benefits in the study area represent significant economic benefits provided by the Upper Mississippi River and are likely to be affected by many of the navigation project alternatives. The Corps should commence with a detailed analysis of impacts on recreation use benefits, going beyond the consideration of only boater safety and lock delays.

Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 61
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 62
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 63
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 64
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 65
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 66
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 67
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 68
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 69
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 70
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 71
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 72
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 73
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 74
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 75
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 76
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 77
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 78
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 79
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 80
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 81
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 82
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 83
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 84
Suggested Citation:"Improving Waterway System Planning." Transportation Research Board and National Research Council. 2001. Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway. Washington, DC: The National Academies Press. doi: 10.17226/10072.
×
Page 85
Next: Summary and Recommendations »
Inland Navigation System Planning: The Upper Mississippi River-Illinois Waterway Get This Book
×
Buy Paperback | $45.00 Buy Ebook | $36.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

In 1988, the U.S. Army Corps of Engineers began an investigation of the benefits and costs of extending several locks on the lower portion of the Upper Mississippi River-Illinois Waterway (UMR-IWW) in order to relieve increasing waterway congestion, particularly for grain moving to New Orleans for export. With passage of the Flood Control Act of 1936, Congress required that the Corps conduct a benefit-cost analysis as part of its water resources project planning; Congress will fund water resources projects only if a project's benefits exceed its costs. As economic analysis generally, and benefit-cost analysis in particular, has become more sophisticated, and as environmental and social considerations and analysis have become more important, Corps planning studies have grown in size and complexity. The difficulty in commensurating market and nonmarket costs and benefits also presents the Corps with a significant challenge. The Corps' analysis of the UMR-IWW has extended over a decade, has cost roughly $50 million, and has involved consultations with other federal agencies, state conservation agencies, and local citizens. The analysis has included many consultants and has produced dozens of reports.

In February 2000, the U.S. Department of Defense (DOD) requested that the National Academies review the Corps' final feasibility report. After discussions and negotiations with DOD, in April 2000 the National Academies launched this review and appointed an expert committee to carry it out.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!