For privacy purposes, this affidavit has been slightly modified from the affidavit filed with the Office of Special Counsel

For privacy purposes, this affidavit has been slightly modified from the affidavit filed with the Office of Special Counsel. Thirteen words have been deleted and two words have been replaced for grammatical consistency. These changes have no material affect on the content.

This affidavit was filed with the Office of Special Counsel, an office that takes disclosures of serious wrongoing from federal employees and is required to assure that agency heads, in this case the Secretary of Defense William Cohen, conducts a thorough investigation.

Affidavit of Donald C. Sweeney

I, Donald C. Sweeney II, under oath, state as follows:

I hold a Ph.D. in economics from Washington University of St. Louis, MO and have been employed as a regional economist with the U.S. Army Corps of Engineers since February 1978. During this time, I received annual performance evaluations, all of which have been excellent, including the most recent performance evaluation covering the rating period ending October 1999. I am providing copies of all those performance evaluations that I currently have in my possession in Appendix 1. I have served as a supervisory economist in the St. Louis District of the Army Corps of Engineers overseeing the work products of other Corps economists. I have received numerous awards for public service during my 22-year career including awards for excellence in navigation system planning and awards for sustained superior performance.
From March 1993 until June of 1998, I served as the technical manager of the economics work group of the Upper Mississippi River, Illinois Waterway Navigation System Feasibility Study. This study is a six year, more than $50 million investigation of the feasibility of making up to $4 billion of improvements to this navigation system that was originally scheduled to complete in 1999 but which will now be complete in 2000. Other members of the work group I directed included economists from three Corps of Engineers District offices and two Corps of Engineers Division offices. From June of 1998 until June of 1999, I served as the advisor to an "economics" panel created by General Phillip Anderson, Commander of the Mississippi River Division, that was subsequently made responsible for the economic analysis in the study although this panel largely ceased to function in October of 1998.
Under the Federal statutes, regulations, and policies that apply to the planning of all Army Corps of Engineers water resource projects, the Federal objective "to contribute to the national economic development consistent with protecting the Nation’s environment, pursuant to national environmental statutes, applicable executive orders, and other Federal planning requirements." This means that for the Corps of Engineers to recommend a project its estimated national economic development benefits must exceed the estimated costs of building and operating the project and the project must be consistent with protecting the nation’s environment. The benefit and cost analysis is a primary responsibility of the economics team and manager on a study. At the time I was removed from responsibility for the economic work products of the study, the study schedule called for the Corps of Engineers to release to the public in September 1998 its recommended plan. At that time, the benefit and cost analysis of the economists working on the study indicated that major, expensive, large-scale, structural changes to the navigation system were not warranted for the foreseeable future. Rather, the benefit and cost analysis indicated that measures the navigation industry could undertake itself or relatively inexpensive small-scale Federal measures were the only measures economically warranted.
The September 1998 date for release of the Corps’ recommended plan and economic and environmental analyses was postponed. Subsequently, in July of 1999, the Corps of Engineers at a series of public workshops announced that a "new" economic benefit and cost analysis showed that immediate, expensive, major structural changes to the navigation system were now justified.
This affidavit describes how the benefit and cost analysis has been intentionally and deliberately altered to produce a seemingly favorable recommendation for immediate large-scale expensive improvements (essentially doubling the length of seven system locks) as a result of the instructions of top Army Corps of Engineers officials. It discusses how new personnel were put in charge of the benefit and cost analysis and explicitly instructed not to produce their best unbiased analysis or, for that matter, any analysis they actually believed to be valid or in compliance with Corps of Engineers regulations, but instead to produce an "analysis" that immediately "justified" large-scale, expensive structural improvements. At first this effort attempted to find a justification for the project that was at least plausible. This attempt is itself contrary to Corps of Engineers Regulation ER 1105-2-100 which requires planning studies to evaluate all Federal water resource projects based on the most likely future project parameters. Eventually, even the self-serving standard of plausibility was abandoned when it failed to support the directed study results, and key inputs to the benefit and cost analysis were arbitrarily altered by top Army Corps of Engineers management to finally produce a seemingly favorable benefit to cost analysis for immediate, expensive, large-scale system investments. These altered benefit and cost numbers had no meaningful professional analysis backing them and were at least in many cases based on clearly erroneous rationales provided or directed by senior management.
To produce this favorable benefit to cost finding, top Army Corps of Engineers officials directed four key changes to the existing study analysis. First, top Corps officials arbitrarily directed the principal economist to ignore most of the potential for towboats to help each other to move through locks more rapidly during periods of high congestion, a practice that is currently used and that can dramatically reduce congestion at minimal increased public cost. Second, a top Corps official directed the principal economist to alter a key term used in estimating the real economic value of reducing barge congestion. Third, a top Corps official directed that an analysis be altered to state that immediate lock extensions would save millions of dollars that would otherwise have to be spent simply rehabilitating the existing locks. These three changes, separately and together, resulted in dramatically increased estimates for the estimated economic benefits of lock extensions. Finally, a top Corps official directed that engineers arbitrarily reduce the estimated construction costs of extending the length of existing locks despite the fact that the original construction cost estimates were approved by an Independent Technical Review of Corps and other experts. This change significantly reduced the estimated cost of lock extensions. In the final analysis, it appears that all four changes are necessary to produce a seemingly favorable benefit to cost ratio for immediately extending the length of existing locks. This affidavit provides a detailed chronology of how these changes were made and provides many supporting documents.

I. BACKGROUND ON THE UPPER MISSISSIPPI RIVER, ILLINOIS WATERWAY NAVIGATION SYSTEM FEASIBILITY STUDY

The Upper Mississippi River and Illinois Waterway Navigation System Feasibility Study area encompasses a navigation system that includes 43 locks located at 37dams (some dams have more than one lock) stretching from Minneapolis and St. Paul MN at the northernmost extent to the confluence of the Mississippi and Ohio Rivers near Cairo, IL at the southernmost extent. The study area itself covers regions of the states of Minnesota, Wisconsin, Iowa, Illinois, and Missouri. The dams create pools in the rivers with sufficient depths to permit navigation to authorized draughts of nine feet. The locks allow transit between the pools created by the dams. All but four of the locks in this navigation system are 600 or less feet long, while most modern towboats push fifteen barges at a time in a tow that extends nearly 1200 feet long. As a result, many tows in the system must routinely break apart into two smaller units to move through the locks in a time-consuming two-step process, adding roughly 45 minutes to an hour for each tow that must undergo this two-step lockage process. At certain times of the year with heavy barge traffic, this results in some traffic congestion at certain locks, with the delays in processing tows increasing the unit costs of water transportation.
The primary reason for undertaking the UMR-IW navigation system study was to investigate the feasibility of Federal and non-Federal measures designed to speed tow traffic through the locks reducing the time needed to process tows, decreasing the levels of congestion in the system, and, therefore, decreasing the unit costs of water transportation. Time currently wasted by tows waiting for service at congested system locks could then be freed up for other productive uses of these expensive assets. For example, a "typical" movement of a barge from an origin in the Upper Mississippi River system to a southern Louisiana destination and back again currently requires approximately a month for the roundtrip. This typical barge roundtrip had approximately a $9.00 per ton cost measured in the 1994 price levels originally used in the study. Some of this roundtrip time is spent waiting for service at system locks due to congestion at the locks. If the congestion could be reduced and barges processed more quickly through the locks then the roundtrip time could be reduced, thereby freeing the barge to produce more movements of goods.
For example, Lock and Dam 25 near Winfield, MO on the Upper Mississippi River system is often a congested lock. In 1999 approximately 35 million tons of traffic passed through this congested lock. Suppose a measure existed that could reduce the congestion at that lock by an average of five hours per tow. If everything else in the system remained the same then the typical roundtrip would now require 10 less hours on average. This represents approximately a 1.5 percent decrease in the time required to move a typical barge from origin to destination and back again or equivalently about a 1.5 percent increase in the productivity of tows and barges.
The fundamental economic question is what value is associated with this increased productivity? A very rough estimate of the value can be generated directly from the existing data. If there are other potential barge users willing to consume this increased productivity at roughly the going rate of $9.00 per ton, then this increased productivity permits the movement of an additional 500 thousand tons through Lock 25 on this system. This would yield an annual economic benefit of approximately $4.5 million computed as the product of the observed willingness to pay, $9.00 per ton, and the increased productivity of the system, 500 thousand tons. If this hypothetical measure cost less than $4.5 million annually then this measure would increase the net economic benefits of the navigation system. These are rough numbers used for example only and do not reflect the actual more sophisticated analysis in the model, but they provide a good starting point for understanding the economic evaluation of potential system measures.
At current traffic levels evidenced on the UMR-IW navigation system, there is broad agreement among economists that expensive, large-scale measures like providing larger locks to process tows are not economically justified. The rationale for the navigation study was the expectation that the demand for barge transportation would grow over time, leading to increased system congestion and increasing barge transportation costs. Consequently, at some time in the future either large-scale or small-scale measures may increase the economic efficiency of the system by reducing the anticipated growing levels of congestion.
The study was and is required by Corps of Engineers regulation to evaluate a full range of alternatives for decreasing system congestion and improving the efficiency of barge transportation. The full range of alternatives to be investigated encompasses Federal and non-Federal measures. In general, the investigated alternatives are grouped into three broad categories: relatively inexpensive small-scale structural measures, such as providing mooring buoys at the locks or extending guidewalls along the lock approaches that only somewhat improve the efficiency of moving barges through a lock; relatively expensive large-scale structural measures, such as doubling the length of some of the 600 foot long locks thereby eliminating the need for two-step lockages; and very low cost operational changes to the existing system, such as scheduling tows to reduce congestion. As the study progressed, the focus of the study narrowed to congestion at the 600 foot long locks in the system closest to St. Louis because traffic levels are greatest in the southern portion of the study area, making these locks most likely to experience increasing delays and congestion.

II. TRAFFIC GROWTH PROJECTIONS

The economic benefits of navigation improvements designed to reduce future congestion depend, of course, on the future levels of congestion without these improvements. The future levels of congestion, in turn, depend on the future growth in demand for barge transportation, so estimating this growth in demand is fundamental to the economic analysis.
As one of my very first duties as technical manager, I contracted with independent, private sector consultants to provide an estimate of the growth in the demand for barge transportation on the Upper Mississippi and Illinois Rivers through the year 2050. These forecasts were deliberately based on the assumption that future congestion was not a problem and that traffic could grow without increasing costs. The goal, in other words, was to estimate traffic growth if the price of barge transportation remained at today’s real level. This forecast of future demand at existing real transportation prices provides a reference point to begin the estimation of the national economic development benefits of potential measures designed to reduce system congestion and decrease future water transportation costs.
Of course, the future is uncertain and I requested the contractor to prepare these forecasts in probabilistic terms. The forecasts were completed by Jack Faucett and Associates and were based on detailed historic data of traffic flows provided to them for 1991 and 1992. The forecasts predicted the most likely growth rate for future traffic in the absence of increased congestion over the 56 year planning horizon would average approximately 1.1 percent per year, with higher growth rates in the early years and lower growth rates in the later years. Most of this predicted traffic growth would result from an increase in grain exports delivered to southern Louisiana ports by barge.
Unlike other parts of the economic analysis, I do not believe any part of this analysis was based on pressure or failed to represent the best judgment of those doing the work. The forecasts were the subject of an intensive technical review conducted by myself, other members of the study team, the North Central Division office of the Corps of Engineers, the Lower Mississippi River Division office of the Corps of Engineers, and the Headquarters office of the Corps of Engineers. The projections of future traffic demands produced by Jack Faucett and Associates were adopted for use by the study without alteration of any kind.
So far, these most likely forecasts of demand for water transportation have proven to be optimistic. Comparing recent observed traffic levels in the UMR-IW navigation system to the most likely traffic forecasts evidences this optimism. For example, Locks and Dam 27 near St. Louis, Missouri is the southernmost lock location on the Upper Mississippi River. Most traffic entering or exiting the Upper Mississippi River goes through these locks. In 1992 Locks 27 processed a total of 81.5 million tons of traffic. By 1999, traffic through this lock assuming continued status quo operation of the system was forecast to grow to approximately 92.0 million tons. This represents a forecast growth of 10.5 million tons in annual traffic. Preliminary Corps of Engineers Lock Performance Monitoring System data indicate that approximately 83.4 million tons of traffic actually passed through Locks 27 in 1999. This represents a real growth of approximately 1.9 million tons during the 1992 through 1999 period. The real observed growth is less than 19 percent of the most likely growth forecast used in the study. Moreover, 1999 was a strong year as traffic levels for the Upper Mississippi River as a whole since 1992 have averaged significantly below 1992 levels. This is evidenced in a graph of historical traffic levels presented by the Army Corps of Engineers study team presented at a meeting of the Governors Liaison Committee to the study on November 18, 1999, which is attached as Appendix 2. This observed low growth relative to the most likely forecast of the study is apparent at all Upper Mississippi River locks and to a lesser extent at the Illinois Waterway locks.
This experience is important because potential water transportation cost savings realized in the near term have a much greater economic value now than potential transportation cost savings realized far off in the future. Low or non-existent traffic growth in the early years relative to the forecast traffic growth used in the study makes the real immediate national economic development value of potential navigation improvements much lower than the values estimated in the study.
Despite this fact, all economic analyses of system improvements, including my own completed in early 1998, use these "most likely" forecast levels of traffic demands. It is important to recognize, therefore, that even the earlier unbiased analyses of the economics work group probably overestimate the real economic benefits of navigation improvements because they use traffic growth projections that, at least so far, appear to be, at best, optimistic.
Further, although this now very likely overestimate of "most likely" traffic growth was unintentional, the interim real observed traffic data should normally result in a revision to these, now optimistic, growth forecasts. This revision has not occurred.

III. BRIEF DESCRIPTION OF KEY FACTORS THAT ULTIMATELY WERE CHANGED TO MAKE LONGER LOCKS APPEAR JUSTIFIED

A. The Willingness to Pay for More Efficient Barge Transportation

The ultimate objective of the national economic development benefit analysis is to estimate the willingness to pay of users for the economic value of reducing the congestion of barge transportation. This willingness to pay for incremental changes in project outputs is the measure of national economic development benefit mandated by Corps of Engineers Regulation 1105-2-100. This Corps of Engineers regulation is itself based on regulations and guidance issued by the Water Resource Council on February 3 and March 10, 1983 signed by President Reagan for all Federal agencies that conduct water and related resource planning.
The relationship between the different quantities of barge transportation, or any other good or service, that consumers are willing to pay for as prices rise or fall is represented by economists as a demand curve. The willingness to pay for any good or service, whether barge transportation or lawnmowers for another example, depends in large part on the alternatives available to potential users of the good or service. When the price of barging, lawnmowers, or any good or service rises, some consumers will continue to purchase as before, other consumers will adjust the quantity they purchase, and still other consumers will take advantage of alternative goods or services.
This demand curve for barge transportation is therefore critical to evaluating the economic benefit of lock improvements designed to reduce future congestion. As system congestion increases, and therefore the price of barge transportation goes up, those who are desirous of moving grain and other goods on the river at existing prices do not have to continue barging the same quantities and can take advantage of other lower cost alternatives available to them. For example, the best analysis of grain markets in Iowa indicates that approximately 17 percent of Iowa grain is transported on the Mississippi River, primarily for export. If barge prices rise in the future, grain merchants and farmers could send more of their grain by railroad to New Orleans for export. They could export larger amounts via rail through the Pacific Northwest. More likely, they could decide not to export as much grain but instead send a little more to local processors, such as ethanol plants or hog feedlots. How much would these consumers of barge transportation be willing to pay for navigation improvements to lower the price of barging? The answer is the difference between that price and their next best alternative, and this difference is the economic value of navigation improvements to them.
Another consequence of the existence of alternatives to barging is that the river will become less congested than otherwise predicted by the most likely traffic growth estimates developed for the UMR-IW study. These forecasts predict the most likely barge traffic growth based on transportation costs and prices remaining constant at existing real levels. But if congestion makes barging more expensive, some of the goods that would otherwise move by barge will go to other destinations or use alternative modes of transportation such as rail or truck and therefore will not contribute to traffic delays. For this reason, even those who continue to use barge transportation despite the higher costs will, in fact, have to face less congestion than implied by the traffic forecasts. In effect, they benefit from the fact that others will decrease their use of the river. This means that determining how much commerce will shift away from the river as barge costs increase is critical to predicting how much congestion will truly develop on the river and therefore how expensive barging will become in the absence of navigation improvements.
A Corps of Engineer Regulation, ER 1105-2-100, and the "Economic and Environmental Guidelines for Water and Related Land Resources Implementation Studies" both require in identical language that all planning studies use the willingness to pay of potential consumers as embodied in a demand curve as the measure of national economic development benefits. In estimating how much additional barge traffic will use a river as a result of a project and how much that increased level of use contributes to the national economic development objective, these regulations require that planners consider all the different alternatives available to those who want to move goods in the absence of the project.
Economists call this analysis of "how much" goods will shift away from the river in response to a price change a question of the elasticity of demand for barge transportation. The more readily grain and other goods will find an alternative to using the Upper Mississippi River as the river becomes more congested and barge prices rise the more elastic is the demand for barge transportation. The less readily grain and other goods will find an alternative to using the Upper Mississippi River as the river becomes more congested and barge prices rise the more inelastic is the demand for barge transportation.
Elasticity depends on the availability of substitutes for water transportation as well as the amount of time producers and consumers of barge transportation have to respond to market changes. Generally, the more close substitutes or alternatives available to consumers and producers the less the willingness to pay for increased levels of barge transportation. Similarly, the more time available to seek alternatives and substitutes the lesser the willingness to pay for increases levels of barge transportation.
Prior to this study, despite the planning regulations, Army Corps of Engineers analyses of inland navigation projects had ignored this fundamental point that those who use water transportation have other alternatives and will adjust their desired level of usage if the price of barge transportation rises. They in effect assumed a perfectly inelastic demand of system users for barge transportation. This assumption violates fundamental principles of economic analysis and can result in an order of magnitude overestimate of the national economic development benefits of a navigation project.
As part of the economic analysis for this project, the economics work team and I developed a model that incorporated for the first time the elasticity of demand for barge transportation in a system model to be used to estimate the navigation-related national economic development benefits of potential system changes to reduce congestion. In the various documents cited later in this affidavit or provided as Appendices, this model is referred to as SEM, ESSENCE, or the St. Louis model. SEM refers to the conceptual economic model. ESSENCE refers to the computer model implementing the SEM. The St. Louis model denotes the SEM, ESSENCE, or both. In the ESSENCE, a parameter termed "N" or sometimes the "N value" technically defines the "shape of the demand curve for individual origin, destination, and commodity specific potential system movements" and plays an important role in estimating economic benefits. The N value for a potential system movement is directly related to the absolute value of the elasticity of demand for that movement. The greater N for a given potential movement, the more elastic is the demand for barge transportation, the better the alternatives to barge transportation appear, and the lower the likely real growth in future barge traffic without system improvements. In other words, the higher the N values for movements, the lower the national economic development benefits of potential expensive, large-scale, navigation system improvements.
In the original ESSENCE model, we estimated the N value for grain and other raw agricultural products, which are by far the most important commodities by volume, to be equal 2.0. All other commodities were given an N value of 1.0. Even assuming the optimistic most likely demand predictions by Jack Faucett and Associates, with these N values, the estimated costs of doubling the length of any subset of system locks or constructing new 1200 foot long chambers any time in the foreseeable future exceeded the estimated benefits substantially. There is legitimate uncertainty regarding the appropriate values of N for agricultural and other products, but as I discuss below, even much lower N values led to the same conclusion.
As described below, this model was subject to repeated independent technical review both within and without the Army Corps of Engineers and was strongly affirmed by each review. In the course of finding a way to justify doubling of the lengths of locks in 1998 and 1999, the N values were one of the key factors subject to manipulation by senior Corps of Engineers officials.

B. Rehabilitation Cost Savings

Another source of national economic development benefits that can be the result of potential structural projects to improve water resource facilities is the postponement or avoidance of rehabilitating or replacing existing components of facilities that would otherwise need rehabilitation or replacement. In this event, these rehabilitation cost savings, if real, are a national economic development benefit of the potential project as those resources that would have been consumed by the rehabilitation are freed up for other productive uses.
As part of navigation study, the engineering and economics work groups evaluated whether there would be rehabilitation cost savings as a result of implementing any of the alternatives to decrease congestion at the locks. Throughout 1997 and early 1998, Jeff Marmorstein and I and others participated in a series of meetings with the engineering work group, including Dennis Lundberg and Bob Hughey and others, that examined in detail the construction cost elements of the extended lock alternatives. At these meetings we were informed that no rehabilitation costs for components of the existing locks were included in the construction cost estimates for the extended locks. For this reason, regardless of what future rehabilitation costs proved to be, there could be no immediate rehabilitation cost savings as a consequence of extending existing locks.
Subsequent to these meetings, Jeff McGrath, an economist in the St. Paul District at my direction, working with members of the engineering work group, completed a detailed analysis of the need for future rehabilitation for the without project condition. The principal reason for this analysis was guidance from Corps of Engineers Headquarters that major rehabilitation expenditures on the Upper Mississippi River must be identified as an integral part of the evaluation of any potential investments on the navigation system. Mr. McGrath’s analysis focused on seventeen major components of the all the existing locks in the navigation system study and concluded that no rehabilitation costs for these components would be necessary with or without extended locks for 35 to 40 years, i.e., until at least 2033. These conclusions were embodied in a report prepared by McGrath in early 1998 and are reflected in the public minutes of an Economic Coordinating Committee meeting on May 11 and 12, 1998 which are attached as Appendix 3.
This analysis is important because it shows that potential rehabilitation cost savings from extending locks would be small even if the costs of extending locks did include significant rehabilitation of existing locks (in contrast to what we were told above). At best, extending the existing locks in the near future would postpone future rehabilitation costs from around 2035 to some period thereafter. The present value of these foregone rehabilitation costs so far off in the future would be very small.
Mr. Lundberg noted correctly at a meeting in early 1998 in the Rock Island District office that I attended that some potential rehabilitation expenses had not been examined by Mr. McGrath’s report. However, Mr. Lundberg also noted these other rehabilitation costs were not expected to change with or without the lock extensions. For that reason any such rehabilitation costs were irrelevant to whether extended locks could save future rehabilitation costs.
The conclusion that there was no real potential rehabilitation cost savings was documented to the economics team in a statement contained in the document entitled "Engineering Hand-off to Economics" dated March 18, 1998. Key pages from this document are attached as Appendix 4. This document contained performance and cost data compiled by the engineering work group for large-scale measures including lock extensions that had undergone extensive Independent Technical Review in late 1997.
As described below, a major change was made to this analysis in May and June 1999 in an effort to attempt to justify immediate large-scale measures.

C. Industry Self-help

One of the means by which tows can themselves reduce the time it takes to proceed through a lock is to receive help from other tows also waiting for lockage. This practice is termed industry self-help. Basically, another line haul towboat waiting at the lock disengages from its barges and assists in removing the first half (called the first "cut") of a locking tow and then removes this first cut to a remote location to await the second cut to remake the original tow. Historical experience developed by the study has shown that tow operators will do this when traffic builds up at locks. The engineering team estimated that "industry self-help" would reduce the time it takes to transit a lock by an average of 18 minutes for up-bound tows and 23 minutes for down-bound tows, or roughly 20 minutes average overall. This conclusion was summarized in the Study’s December, 1997 newsletter, attached as Appendix 5. The March 18, 1998 revised Engineering Hand-Off to Economics (attached previously as Appendix 4) reduced slightly the estimate of the time savings from increased usage of industry self-help and increased the costs by requiring that DeLong Piers be installed to extend the existing upstream guidewalls of the locks for safety reasons.
This practice has important consequences for estimating the benefits of larger locks. Corps of Engineers regulation ER-1105-2-100 requires this practice to be explicitly included in the without project condition. This practice helps insure the most efficient use of the existing system. Larger locks, according to the engineering team, reduce transit times per lock by approximately 55 minutes (depending on the lock). But when tow operators can themselves significantly reduce transit times by up to 23 minutes under certain conditions, the net reduction in lockage time attributable to longer locks is reduced when measured against the without project condition.
Although industry self-help occurs today, the addition of nearby mooring cells would further facilitate the practice. As the 1997 newsletter reveals, one possibility considered was to construct six mooring cells at each lock at an estimated cost of $1.1 million per lock. However implemented, Army Corps of Engineer regulations and the Principles and Guidelines require that self-help be treated as an alternative to extended locks.
The ESSENCE model therefore incorporated a technique for estimating when tows would provide each other self-help based upon whenever the congestion at a lock reached a certain level, 12 tows waiting for lockage. The model then reduced the lockage time for those double lockages that could benefit from self-help in the without project condition.
The model, depending on the precise lock in question, indicated that the tow industry could use self-help at the locks in the future for the majority of double lockages. Because each use would reduce transit time by roughly 20 minutes even without extended locks, this estimate of real world conditions significantly and dramatically reduced the estimated economic benefit of extended locks. As described below, as late as November 1998, the economic analysis indicated that self-help alone would fully address the future transportation needs of the system.
As described below on January 14, 1999, at a navigation team leaders meeting in the Rock Island District office it was announced that officials ordered an important restriction to be used in the model limiting the percentage of time barge companies would use industry self-help.

D. Cost Estimates for Extended Locks

In any benefit and cost analysis, the estimated costs of a project obviously play a critical role. In the course of the economics study while I was still leading the economics work group, cost estimates for all the large-scale lock expansions dramatically decreased. Original cost estimates using traditional lock construction techniques put the cost at approximately $380 million per lock or $1.9 billion for five locks. These original cost estimates were produced for an earlier study and just served as a starting point for the investigation of this study.
The engineering work group recognized early that such a high construction cost was likely not to be cost-effective for addressing future congestion. It therefore investigated and developed an innovative construction technique that would keep the existing locks in place and simply extend them using a great deal of pre-fabricated construction. For example, unlike traditional Army Corps construction techniques, the proposed new lock extension construction would not require the construction of a costly temporary (coffer) dam. These new techniques, reflected in engineering alternative 2R reduced the construction cost estimates roughly in half.
One part of any Corps of Engineers cost estimate includes a "contingency" factor for potential cost overruns related to unforeseen conditions encountered during construction. This contingency is included because of experience that costs rarely turn out to be lower than predicted and often turn out to be higher. Typically Corps projects include contingency factors of 25 percent. That typical 25 percent contingency cost was increased in this study to 35 percent in response to the Independent Technical Review of the Engineering Appendix to the study and expert comments received by the engineering work group because the lower construction cost estimates relied on novel construction techniques. This 35 percent contingency cost was reflected in all large-scale cost estimates transmitted to the economics work group for use in the benefit/cost analysis as documented in the Engineering Hand-Off to Economics revised March 18, 1998 (previously attached as Appendix 4).
As described below, this contingency factor was suddenly reduced to 25% in late May and early June 1999.

IV. DESCRIPTION OF EVENTS SINCE 1997

A. October 1997 through May 1998

On October 29, 1997, in response to my earlier request, I briefed the economic methodology the economic study team had developed for the study to a large group of invited Corps economists and planners in St. Louis. All Corps economists and planners with an interest in navigation studies were invited and asked to comment and to review on the UMR-IW economic methodology and model.
Oral feedback from the briefing was extremely positive. On December 5, 1997 I also received an email from Jesse McDonald, an economist at the Mississippi River Division, including copies of the comments of several participants. These comments are attached as Appendix 6. The comments of Robert Daniels, then the chief economist of the Army Corps of Engineers at Headquarters, stated: "My simple answer is that it looks like the greatest advance we have made in moving theory to practice in quite a few years. What does it mean to the future analysis? Well, I’d say it means that we will have more correct and therefore more supportable results." Steve Cone, another economist in Headquarters, described it this way: "An appropriate move toward what has been viewed as theoretical (but only because models have not been developed to measure what we know is true). I hope we continue to move toward models that better measure willingness to pay."
In November and December of 1997, Jeff Marmorstein and I used the model to generate preliminary estimates of the economic benefits of various alternative improvements for the Upper Mississippi navigation system. These preliminary analyses concluded that industry self-help alone might fully address future transportation needs and at most only small-scale, low cost system improvements appeared justified for the foreseeable future. The estimates of the economic benefits also indicated that large-scale improvements would not be justified at any time in the 50-year study period.
At my request in late 1997, Mr. Richard Manguno an economist of the New Orleans District office was designated as the independent technical review manager for the products of the economics study work group. Army Corps of Engineers regulations require an independent technical review of the work products of feasibility studies.
On February 3-4, 1998, in Vicksburg MS, I first described the preliminary results of the economic benefit and cost analysis at a public meeting of the environmental work group of the UMR-IW study. I presented the results as subject to change and preliminary, however, I did communicate that only small-scale measures appeared warranted for the foreseeable future.
These preliminary working results generated a few newspaper articles stating that the "Corps of Engineers rejects rebuilding lock system." An article appeared in a newsletter put out by American Rivers, an environmental group, and I believe later some other newspapers carried the story, including perhaps the St. Louis Post Dispatch and the Des Moines Register.
On March 11, 1998, I formally forwarded to Rich Manguno of the New Orleans District office documentation to begin the required Independent Technical Review (ITR) of the work products of the economics work group. As part of the ITR, Mr. Manguno would pick a number of economists, both within and outside the Army Corps of Engineers, to evaluate the work products. The St. Louis District is under the command and control of the Mississippi River Valley Division (MVD), which has its headquarters in Vicksburg, Mississippi. As part of its command and control duties the MVD is required by regulation to assure the ITR of all work products used in feasibility studies under its command and control.
In early April 1998, the entire study team and I briefed the MVD commander and his staff on the ITR process for the entire study and other issues. At this briefing I formally told the commander and his staff that this study would likely not produce economic justification for large-scale measures at any time in the foreseeable future. I explained how previous Corps of Engineers economic evaluations of navigation system improvements had overstated project benefits. The commander and his staff, including Don Herndon and George "Dusty" Rhodes, expressed concern with our conclusions.
Minutes of an April 7, 1998 Navigation Study Team meeting (Appendix 7) indicate that the schedule at that time called for releasing the initial NED analysis, i.e., the benefit/cost analysis, in April and presenting the NED and alternative plans in July. These minutes also note, "Based on the preliminary, initial NED analysis and its assumptions, industry self help appears to be able to meet the need for system transportation improvements within the study horizon." There was also discussion of the estimates of rehabilitation costs, and the minutes include my uncontradicted statement "that this analysis should not significantly effect the outcome of the study, since it essentially impacts the with and without project condition in the same way."
On April 23, 1998 at Fort Leonard Wood, Missouri, in response to these concerns, I briefed General Fuhrman (the Director of Civil Works for the Corps of Engineers), General Van Winkle (the Commander of the Lakes and Rivers Division (LRD)), General Anderson (the Commander of the MVD), Colonel Mudd (the Commander of the Rock Island District), and Colonel Hodgini (the Commander of the St. Louis District. Other staff and study team members were present at the briefing. In the course of doing so, I addressed why this analysis differed from previous Corps of Engineers navigation analyses. I explained how the information we had developed on the origins and prices paid to producers of Iowa grain that currently moved on the Mississippi River showed that farmers utilized abundant alternative destinations for their grain, which meant that they must be responsive to price changes in barge transportation. (Overheads from this presentation are also included in Appendix 7.)
None of the participants expressed any objections to the analysis. General Fuhrman decided to undertake a process to evaluate the potential implications of the UMR-IW results and models. He stated that he did not want to revisit the economics of past projects, but wanted to ensure consistency in evaluation of future projects among all Corps offices undertaking such studies. The meeting concluded with an understanding that General Fuhrman’s staff would undertake an investigation to assure the consistency of economic analyses amongst all divisions and districts conducting these analyses. Steve Stockton, Chief of Planning at Corps Headquarters, was assigned to this task initially, but to my knowledge this investigation was never completed.
On May 27, 1998, I attended a meeting in Huntington, WV to discuss comments received from economists at LRD as part of the Independent Technical Review regarding the economic model and data used in the study. The feedback from this meeting was positive although the economists at LRD withheld judgment on some of the details of the model pending further review. This outcome was described by Dudley Hanson, at that time the overall project manager of the study in a June 2, 3, 1998 meeting of all the technical managers of the study in Bettendorf, Iowa, which is recorded in a memorandum dated June 4, 1998 from Bradley Thompson to the participants in that meeting (Appendix 8). As this memorandum describes, the LRD economists wanted a users manual for the model to explore some of its details.
The minutes of the June 2,3, 1998 technical managers meeting also provide some discussion of self-help. Denny Lundberg, an engineer in the Rock Island District, stated that the navigation industry had just "come to understand the importance of this measure," opposed it, and was "not willing to talk about this measure any further." But Lundberg also reported that the self-help analysis of the engineering work group had already passed Independent Technical Review.
To further explain the details of the economics model, a follow-up workshop was conducted in St. Louis on June 10, 1998 for economists participating in the ongoing ITR and others. Jeff Marmorstein led the workshop. The conclusions and recommendations of the workshop were that the concepts underlying the model were sound. The workshop participants recommended that it might be necessary to do more work to estimate the "elasticity of demand" for barge transportation, i.e., to determine the appropriate N values for use in the ESSENCE model. I agreed with that recommendation. A memorandum from me to the members of the ITR economics team summarizing these conclusions dated June 16, 1998 is attached as Appendix 9.
In short, as of June 10, 1998, the economic analysis that showed that immediate lock extensions and other large-scale measures were not justified had undergone both informal and formal review, and the only recommendation was the possible need for further data checking of the N values. I agreed with that recommendation.

B. June 11 1998 through April 1999

On June 13, 1998, General Anderson received a briefing in St. Louis on the status of the ITR of the economic analysis. Col. Hodgini, Dudley Hanson, George "Dusty" Rhodes (Chief of Planning Directorate at MVD), Dean Caldwell (Executive Assistant to the MVD Commander) and Rich Manguno were present at the briefing. I was originally supposed to attend the meeting but was told shortly prior to the meeting by Owen Dutt, my supervisor, that he and I should not attend the briefing. My understanding of the meeting is based on descriptions given to me by Colonel Hodgini and Mr. Manguno orally and on a memorandum dated June 15, 1998 (Appendix 10).
Mr. Manguno had pre-briefed Mr. Rhodes that the ITR was going to conclude that the model was correct and suitable for use in evaluating the economic impacts of alternative changes to the UMR-IW navigation system. Mr. Rhodes stated that he would challenge the conclusion of the ITR. He directed prior to the briefing of General Anderson that alternatives be developed to alter the management structure of the study team. He directed that these alternatives include the replacement of me by a panel of MVD economists chaired by the project manager Dudley Hanson, an engineer, and he stated he was going to recommend this alternative to General Anderson.
At the briefing for General Anderson, Mr. Manguno reported the status and likely conclusions of the ITR. Mr. Rhodes, who is not an economist, presented his recommendation to transfer control of the economics work by stating that I was on a mission "out to shut down the Corps." The briefing concluded with the understanding that General Anderson would consult with General Furman and make a command decision on what to do.
On June 17, 1998, General Anderson sent a memorandum to the Commanders of three of his districts. This memorandum (Appendix 11) instructed the following:

Dudley Hanson would be appointed the project manager for the overall study as his sole responsibility. Previously, Mr. Hanson acted as the supervisor for the project manager of the time, Mark Gmitro.

Control over all economic analysis was transferred to the newly created panel consisting of the Chiefs of Economics in three MVD Districts, Mr. Rich Manguno, Diane Karnish, Paul Soyke, and a representative from the Navigation Studies Center in LRD, Wes Walker, who was already participating in the ongoing ITR of the economics work products.

The panel was directed to produce all necessary economic analyses and to identify the alternative that maximized the net economic benefits within 90 days.

I was named an advisor to the panel. Later, the Corps of Engineers Headquarters office also appointed an observer to the panel, Mr. Ron Conner.

On June 22, Paul Soyke sent a memorandum, at the "direction of the Project Manager," Dudley Hanson, describing the panel’s roles (Appendix 12). It stated that Dudley Hanson is the chairman of the panel. Dudley Hanson is not an economist. It also instructs the panel members that there are 17 key assumptions described in my write-up of the SEM economics model that the panel will review.
The economics panel held its first face-to-face meeting July 1-2, 1998 in New Orleans. I did not attend this meeting. A CEMVR-PD-C memorandum dated July 6, 1998 (Appendix 13) documents this meeting. The panel evaluated the 17 assumptions underlying the economics model and concluded, "there were no major problems or issues with the assumptions."
At this meeting, Dudley Hanson also directed that the panel contract with Dr. Mike Bronzini of Oak Ridge National Laboratories, to evaluate the ESSENCE economic model using another model to determine if both models arrived at the same results under the same assumptions.
On July 23, 1998, the economics panel met for a second time. Its proceedings are reflected in a memorandum prepared by Ron Conner dated July 25, 1998 (Appendix 14). Dr. Bronzini discussed his progress. I urged that the panel devote its full energy to improving information about the likely shape of demand curves. The panel agreed to organize an "expert elicitation panel" of outside agricultural economists to provide advice on the elasticity of demand for water transportation for agricultural products.
On August 6 and 7, 1998, the expert elicitation panel of outside experts met to help the economics panel determine the characteristics of the demand for water transportation of agricultural products. The meeting was held in Chicago, IL with the following experts in attendance: Harold Hommes, Iowa Dept. of Agriculture; Brad Clow, Sparks Companies Inc.; John Bitzan, North Dakota State University; Steve Fuller, Texas A&M University; and Jerry Fruin, University of Minnesota. Corps of Engineer attendees included Paul Soyke, Ron Conner, Mark Hammond, and myself. The meeting is documented in two memorandums, CECW-PD (Conner) dated August 10, 1999 and CELRH-NC also dated August 10, 1998, both of which are attached as Appendix 15.
As these memoranda indicate, the panel agreed with the basic (downward sloping) shape of the demand curve in the ESSENCE economic model. It recommended that more data be gathered to determine the precise shape of the curve for various origins and destinations for grain products, but it believed that the appropriate N value for the demand curves for raw agricultural products was probably somewhere between 1 and 2.
On August 14, 1998, the Independent Technical Review of the economic model and other work products was finalized in a memorandum (Appendix 16) transmitted through Owen Dutt (Chief of the Planning Division in the St. Louis District) to Dudley Hanson. It stated that the "economic concepts underlying" the economic analysis are sound. It recommended additional efforts to define the shape of the demand curves for water transportation, i.e., to determine the elasticity of demand for water transportation.
On August 22-23, 1998, the economic panel met for the third and final time in Rock Island. The meeting is documented in a memorandum from CECW-PD (Conner) dated August 26, 1998 (Appendix 17). In this meeting the economics panel recommended continuing with the model developed for the study by the study team with a "compromise" demand curve for water transportation of all raw agricultural products movements characterized with an N value of 1.5. For non-agricultural products the economics panel recommended an N value of 1.0.
The memorandum from General Anderson had assigned the economic panel the responsibility of producing a preliminary NED plan by September 17, 1998. I do not know precisely who made the decision, but at some point in September, Dudley Hanson directed Rich Manguno to analyze various selected alternatives (combinations of small and large scale measures at selected system locks) through the model using different selected values for N and different growth predictions for future traffic. (As discussed above, the Jack Faucett analysis had provided a high, medium and low estimate for future traffic growth.) I have an email dated September 3, 1998 from Paul Soyke to the economics panel members and others indicating that there had been a teleconference the previous day, the outcome of which appears from this email to have been a decision not to produce a single recommended alternative. Instead, a range of alternative scenarios would be analyzed. "The team should then determine an alternative, based on these results, that appears to be the most likely to justify large scale improvements in the near future." (Appendix 18).
Rich Manguno came to the St. Louis District on September 14,1998 and stayed through September 18, 1998 to run the various alternatives through the model using various different assumptions for N and growth predictions. He had been instructed by Dudley Hanson to examine how these different assumptions affected the results. Jeff Marmorstein and I assisted him. We concluded that any reasonable values for either the N parameters or the growth rates left unaltered the conclusion that no large-scale improvements were justified for at least a long period.
On September 18, 1998, I was called to a private one on one meeting with the St. Louis District’s Deputy for Project Management, Mr. Gerald Barnes. He informed me he had had conversations with MVD staff, specifically Dusty Rhodes and Don Herndon. He told me to find a way to justify large-scale measures in the near term for the UMR-IW navigation system or the MVD office would find an economist who would and I would be out of my job as technical manager of the economics work group. As I remember, he asked if I had a family to support or words to that effect. I told him I could not find such a justification because it was beyond the bounds of reasonableness given the economic data and model runs to date. I described this meeting to Mr. Manguno immediately on walking out of the meeting.
On September 23, 1998, General Anderson, Colonel Mudd and members of their staffs briefed General Furman and key members of his staff. I do not know who attended this briefing, as I was not invited to attend.
On the same day, there was an exchange of emails between William Arnold, a planner with the MVD, and Dusty Rhodes, the Chief of Planning of MVD (Appendix 19). In these emails, Mr. Arnold communicates that they have been trying to choose scenarios out of the variety that could be picked to justify the project that are "reasonably plausible," but warns that they could end up, without care, putting together assumptions that together could not meet this standard. Rhodes writes back that they should stick "with the concept of scenarios that are reasonably plausible." Army Corps of Engineers regulation ER 1105-2-100 and the Principles and Guidelines require that scenarios be chosen that are the "most likely condition expected to exist in the future."
On September 25, 1998, Dudley Hanson sent an email to Mr. Manguno with copies to many individuals involved in the study including myself (Appendix 20). This memorandum instructed Rich on the "task that you, as leader of the economics work group . . . have before you." The memorandum includes a description "Briefing of MG Fuhrman and staff: Guidance Received" and states in part as follows: "We, the Corps, are the Federal Government’s advocates for inland waterways. There is a need to improve the system. The well being of the Midwest depends on agricultural exports. We need to figure out what the demand curves mean. If the demand curves, traffic growth projections, and associated variables that the economics model can consider, do not capture the need for navigation improvements, then we have to figure out some other way to do it . . . . MG Fuhrman will approve a schedule change. By the end of three weeks . . . he directs that we develop evidence or data to support a defensible set of capacity enhancement projects. We need to know what the mechanism is that drives the benefits up. The rationale should err on the high side." If there is any ambiguity as to the meaning of this instruction, Dudley Hanson includes the instruction: "In anticipation of the need [for additional help for Manguno] I have asked folks to work on the rationale that relates directed enhancements to plausible world conditions that will underlie the need for these enhancements, to be developed qualitatively and then quantitatively using our analytical tools."
On the same day, Dudley Hanson also sent a memorandum to a large number of people involved in the study (Appendix 21), including myself stating: "MG Fuhrman now tells us that we are the advocates for inland navigation, which sounds to me to be a distinctly more proactive posture than what I've always pictured our role. I've thought of us as "stewards" of inland navigation, i.e., we execute public policy regarding improvements and level of service, and we grease the machinery and repair it. This overt advocacy role, to me, is a new departure. We'll have to work on a story line, and in fact, that's one of the things we'll be doing over the next few weeks. I, for one, have no problem with this, provided our chain of command supports, and helps explain, this new view when the heat comes up. It's pretty clear to me from Wednesday's meeting in Washington that the support will be there when it's needed."
Mr. Manguno has informed me that he expressed concerns with this instruction in writing to Dudley Hanson.
On October 2, 1998, in response to these concerns, Colonel Mudd sent Mr. Manguno an additional memorandum (Appendix 22). Among other things, this memorandum states: "MG Fuhrman has clearly stated that the Corps has the responsibility as the Federal Government's advocate for the inland waterway system. To help in the execution of this responsibility, you will develop the economic component of the case for a recommendation that includes near term improvements, recognizing that the nation is better served by improvements that err on the large-scale side than by actions that err on the underdeveloped side. The case will be based on explicit considerations of our position in the world with respect to competitiveness and reliability. Considerations may also such include [sic] items as national policy regarding grain exports, balance of trade considerations, and risk and uncertainty. Making this case will be the responsibility of the PM/Formulation group, with the help and input of the Economics Panel."
Also on September 25, 1998, Gerald Barnes sent an email to Colonel Hodgini, Diane Karnish and myself reiterating some remarks by General Fuhrman to the navigation industry stating, among other things, that, "HQ will be ‘decisively engaged’ in the results of the Navigation Study." He ends the memo with the line, "It’s pretty clear to me where this is heading." (Appendix 23)
In a memorandum dated September 27, 1998, Major General Anderson extended the Economics Panel for the duration of the feasibility study (Appendix 24). The panel was made responsible for all the economics work products in the study. Rich Manguno replaced Dudley Hanson as panel chair. Col. Mudd was given ultimate control of the panel. The panel was made responsible for production of all economic products for the navigation study.
Despite the above instruction, the panel never met again nor was consulted with in any way until I sent a memorandum by e-mail to Col. Mudd in June 1999 reminding him of the panel’s responsibilities. These events are described below.
On November 24, 1998 a document entitled Upper-Mississippi River- Illinois Waterway System Navigation Study: Preliminary Economics Findings Released to the Public was distributed to the public (Appendix 25). The preliminary findings set forth the benefits and costs of various alternatives under selected alternative assumptions for traffic growth and N values. This analysis left undetermined the system environmental costs to be included in the benefit and cost analysis. No scenario was analyzed using an N of 2.0 for grain. The highest assumed N value estimate employed in this analysis was N of 1.5 for grain. This was the N value recommended for use by the economics panel. Assuming that N value, even without including system wide environmental costs, the analysis showed that extending locks was not justified at any time over the next fifty years.
Subsequent to this public release of preliminary findings in November 1998, Mr. Manguno contracted with Dr. Mark Burton, an economist at Marshall University, to produce additional analysis of the elasticity of demand of commodities that move on the river and to further examine the methodology utilized by the system economic model. Dr. Burton is a recognized expert in the field of transportation economics.
At a January 14-15, 1999 Navigation Study Team meeting in Rock Island, Richard Manguno is told to restrict the model to assume that industry self-help can only be used a maximum of 5% of the lockages, compared to a more than 50% use at congested locks predicted by the model itself. According to the minutes of this meeting (Appendix 26), because concerns have been raised that self-help might impact the environment without additional facilities (presumably mooring buoys), self-help should actually be considered a "with project" option. That change by itself would still preserve self-help as a cost-effective option to extended locks. But, the minutes then state, "Bob Hughey and Brad Thompson highlighted that in detailed coordination with the industry and Corps operations and engineering staff that remote remake facilities will not be provided as a federal action." Bob Hughey is an engineer in the St. Louis District and Brad Thompson is a planner in the Rock Island District.
No other rationale has ever been provided to the economists for this arbitrary restriction on the use of self-help. Currently, tows do in fact use self-help when congestion is significant and it is in the tow operators’ interests to do so.
On March 3, 1999, Dr. Mark Burton submitted his report (excerpts are attached as Appendix 41). His independent assessment reviewed the overall economic model and supported it. He did some empirical analysis, using railroad transportation price and quantity data, to estimate the short-run elasticity of demand for rail transportation of goods other than grain. These estimates of short-run rail price elasticity could legitimately be used to determine the most inelastic limit for these goods, but his report cautions that these are short-run elasticity estimates and they likely understate the real, long run values. The reason is that in the very short-run, those who use the river to transport goods have relatively fewer alternatives, but in the long run, they have more opportunity to find alternatives for water transportation. Dr. Burton also concludes that the best estimate of N for grain is 2.0, rather than the 1.5 that had been recommended by the economics panel.
Chris Brescia was and is the Executive Director of MARC 2000, an association formed by large water transportation companies and others interested in the navigation system to push for large-scale structural improvements on the Upper Mississippi River system. Email documents indicate that there was some discussion in March between Mr. Brescia and at least Colonel Mudd about possibly recommending to Congress major structural improvements even though they did not pass an economic benefit and cost analysis. In an e-mail dated March 11, 1999 from Brescia to Colonel Mudd, Mr. Brescia rejects this idea because he quotes a Corps of Engineers guidance document stating that potential projects without a positive benefit to cost ratio will not receive federal funding (Appendix 27). Brescia concludes in his message, "We will lose on all counts."
By April 1999, Rich Manguno had performed many additional model runs to evaluate the different alternative improvements to the navigation system. As Mr. Manguno had been instructed to do in January 1999, these analyses arbitrarily restricted the use of industry self-help in the without project condition to 5% of the total lockages at each system lock. These analyses also employed lower N values for commodities other than grain computed from Dr. Burton’s short-run rail elasticity limiting values even though Mr. Manguno knew and agreed with Dr. Burton that these were short-term and therefore likely underestimates. These analyses did not incorporate any system-wide environmental mitigation costs. These analyses also used the N value of 1.5 for grain movements recommended by the economic panel even though Dr. Burton had recommended 2.0. Despite all these changes designed to produce the most favorable benefit to cost analysis possible for large-scale navigation improvements, the analyses still indicated that large-scale structural measures, such as extensions of the existing locks, were not justified for any locks in the foreseeable future. Only small-scale measures appeared to be justified. Mr. Manguno took this as strong evidence of the robustness of the earlier conclusions that large-scale measures where not economically warranted in the foreseeable future.
Mr. Manguno reported these results at a meeting of navigation study team leaders on April 19, 1999 in Rosemont, Illinois, which is summarized in draft minutes sent out by Bradley Thompson on April 20, 1999 (Appendix 28). The minutes state: "Rich stated that based on available data the Preliminary NED plan [most economically favorable plan] appears to be five mooring buoys/cells followed by guidewall extensions with powered kevels at UMR locks 20-25." The minutes also note his statement that no improvements "appear justified" at all any locks on the Illinois Waterway.
In short, between June 10^th of 1998 and through April of 1999, all or parts of the economic analysis had been subjected to further review by an economic panel, an independent modeling effort, an expert elicitation panel, an outside economist and a new principal economist. The economist had been instructed by the highest Army Corps authorities to find a rationale for the project. The most cost-effective means of reducing congestion, industry self-help, had been largely eliminated from the study without supporting technical analysis. And Mr. Manguno, the new principal economist, had used N values to run the analyses that were the most favorable for large-scale improvements for which he could find a rationale despite his own reservations. Even so, lock extensions and other large-scale improvements continued to remain unjustified.

C. Developments from May 1999 through the Summer

As part of the study, the Governors of the five states in the study area had each appointed a representative to the Governors Liaison Committee. At the time, a GLC meeting was scheduled for May 18, 1999 in Minneapolis-St. Paul MN at which the study team was to report the preliminary NED plan. The NED plan is that plan that reasonably maximizes the net contributions to the nation’s economy of all potential plans for system improvements, i.e., it is the plan with the highest net economic benefits. It is a special plan required by Corps of Engineers regulation ER 1105-2-100.
A meeting of the Economic Coordinating Committee was also scheduled for May 5-6, 1999. The ECC was formed at the request of the GLC to facilitate the exchange of information regarding the economic aspects of the study. Each state appointed a member to this committee and there were other committee members including Mr. Brescia, a representative from the U.S. Department of Agriculture, and a representative form the Maritime Administration. The ECC held meetings open to the public usually in conjunction with scheduled GLC meetings. The study team expected to use this meeting to explain the economic and other analyses to these representatives so that they could better explain them to their representatives to the GLC.
Around mid-April 1999, Chris Brescia requested a so-called "Economic Summit" that would include top Army Corps officials, some members of the economic study team, and a range of industry representatives to discuss the preliminary results. He requested that this meeting occur on May 5 or 6, 1999. This request is reflected in a memorandum from Dudley Hanson to various team members dated April 14, 1999 (Appendix 29) in which he states: " I think that if MG Anderson grants Chris’ request, it would automatically postpone the ECC. I believe that’s probably one of Chris’ purposes."
On April 15, 1999, in a conference call between team leaders, Dusty Rhodes directed the study team to postpone the ECC meeting and not to make any "additional data releases" prior to the summit. This phone call is documented in a memorandum from Bradley Thompson to many study team participants dated that April 15, 1999. (Appendix 29)
The so-called Economic Summit occurred on May 5, 1999.The meeting included the following representatives from the navigation industry: Norb Whitlock, the head of American Commercial Barge Lines (one of the largest barging companies); Stephen Sheridan from ConAgra/Peavey, one of the largest grain shippers and barge operators, Gerry Brown, head of barging for Cargill; Chris Brescia, Harry Cook, head of the National Waterways Conference, an association of waterway interests, Barry Palmer, Executive Director of DINAMO, Chris Brescia, Executive Director of MARC 2000, and Paul Bertels, with the National Corn Growers Association. The meeting also included several consultants hired by Mr. Brescia, Mr. Sandor Toth and Dr. Jake Haulk. From the Army Corps of Engineers, the meeting included members of the study team, General Anderson, and Colonel Mudd. It also included representatives from LRD and Army Corps headquarters. The meeting was not open to the public, and I am not aware of any communications to the public prior to the meeting indicating it was going to take place. My understanding of the meeting and pre-meeting among Corps staff only is based on conversations with Mr. Jeff Marmorstein, who attended the meeting, as well as some memos I have reviewed describing the meeting (Appendix 30).
The Corps’ pre-meeting meeting was opened by Mr. Don Herndon, Chief of Programs and Project management at MVD, when he informed the Corps’ attendees that their job was to find a way to justify large-scale projects for the UMR-IW navigation system.
Much of the pre-meeting and meeting focused on the central question of the estimates of demand elasticity and appropriate N values. During the post-meeting, Mr. Manguno was asked repeatedly by Dusty Rhodes to identify the N value for grain that would be necessary to show that extending locks was justified in the near term. Rich did not reveal the value, but Colonel Mudd stated that he himself had figured out it would take an n value of about 1.2.
The central outcome of this meeting with industry representatives was that a smaller meeting would be held in Chicago, IL on May 11-12, 1999 between selected representatives of industry, selected members of the study team, and selected LRD representatives. The purpose of this meeting was to come to a meeting of the minds between Corps of Engineers management, analysts, and barge industry representatives. This meeting was also not open to the public.
On May 11 and 12, 1999, the follow-up meeting between industry and the Army Corps of Engineers occurred in Chicago. The meeting included Mr. Brescia, Mr. Sandor Toth, a consultant hired by Mr. Brescia, Dr. Jake Haulk, another consultant for Mr. Brescia, Colonel Mudd, project manager Gary Loss, economist Rich Manguno, and operations research analyst Jeff Marmorstein. It also included Ron Keeney, the Chief of Programs and Project Management of the Huntington District office, from LRD, who was brought in to facilitate and Paul Hanley, an LRD economist. My understanding of this meeting is based on conversations with Mr. Marmorstein and on two summaries of the meeting: a summary by Gary Loss dated June 1, 1999 and a memorandum from Colonel Mudd to General Anderson, which was subsequently forwarded to General Fuhrman with copies to several other participants (Appendix 31).
After Mr. Loss briefed the meeting participants with the current estimates of benefits and costs Mr. Ron Keeney opened the meeting by stating that the study only needed to close the "small" gap between benefits and costs to justify the desired large-scale projects. A small tweak in the benefit estimate and small decrease in the cost estimate would accomplish the closing of the gap. In particular, Ron Keeney stated that an important category of project benefits in LRD is large cost savings from avoiding extensive rehabilitation of old projects. Mr. Keeney noted that the study analysis to date indicated that there would not be any of these rehabilitation cost savings as the engineering work group had indicated that scheduled rehabilitations would be the same with and without potential projects. Mr. Keeney indicated that a reanalysis of this benefit category could be useful in closing the benefit to cost gap.
Colonel Mudd sent a memorandum to General Anderson dated May 14, 1999 (Appendix 31) describing the meeting and its results, which General Anderson forwarded to General Fuhrman. Colonel Mudd stated that "[t]he meeting outcome was a lot better than I had expected." The memorandum states that MVD had agreed to "re-look and adjust were appropriate" the following items. They included cutting the contingency cost estimate from 35 percent back to 25 to reduce the estimated costs of the lock extensions and to reevaluate rehabilitation cost savings.
The minutes of this meeting also describe an agreement that Jeff Marmorstein would do some analysis with Sandor Toth, MARC 2000’s consultant, to determine if Mr. Toth’s analysis could produce an acceptable, lower estimate of the N value for grain. But, as a June 2, 1999 memorandum from Gary Loss to General Anderson, an industry representative and other top Corps officials stated, the "study team indicates that they do not see a defensable (sic) rationale for "Toth'’ approach (Appendix 32).
In a May 28, 1999 Study Update authored by Gary Loss, then the project manager, and broadly distributed, he states that engineering review has resulted in cutting the contingency cost percentage from 35% to 25% and has found substantial rehabilitation cost savings (Appendix 33). This memorandum indicates that the many years of independently technically reviewed prior analysis on both these issues was overruled by a new re-look within two weeks of the Chicago meeting with MARC 2000 representatives and LRD representatives. To my knowledge, no significant documentation of this new "analysis" has ever been disseminated within the navigation study team or made available to the public, and none of this reanalysis has ever been subjected to Independent Technical Review as required by Corps of Engineers regulations.
I have since seen a three page document that I understand was provided to the public explaining this new major rehabilitation conclusion (Appendix 34). It provides no indication of the prior different conclusion. Its rationale appears to be that a document has "defined" major rehabilitation as something that occurs "with an expected extended life of 25 years." Because some locks at issue were last rehabilitated in 1990, this means that major rehabilitation would have to occur in 2015 (in contradiction of Mr. McGrath’s detailed analysis). It further asserts, in contrast to what I was repeatedly told in 1997 and 1998, that the cost of the major rehabilitation items to make the existing locks an integral part of the proposed lock extensions "is included in the estimates for the lock extension alternatives," and thus would result in rehabilitation cost savings in the future.
On May 19, 1999, study team leaders met in Minneapolis, St. Paul MN after the May 18, 1999 GLC meeting at which the Corps was scheduled to release the preliminary economic findings but did not. At this meeting, Rich Manguno was urged one last time to reevaluate the 1.5 N value that he recommended for grain products. He again stated that 1.5 was as far as he could go.
On May 27,1999, Colonel Mudd called Mr. Manguno. Gary Loss, who had by now replaced Dudley Hanson as the overall study manager, was also on the call. In this phone call, Colonel Mudd instructed Mr. Manguno that he should use an N value for grain of 1.2. He further stated that he had developed a rationale for how it could be 1.2. He explained the rationale (later repeated publicly) as follows. The origins of Iowa grain that comes to the Mississippi River can be separated into three regions: a region near the river, a middle region of Iowa, and a farthest from the river region. The expert elicitation panel suggested that grain produced near the river had a relatively lower N value and grain farthest away a relatively higher N value. Col. Mudd stated that an N value of 1.2 could be justified by taking a weighted average of 1.0, 1.5, and 2.0 using as weights the proportion of grain delivered to the river from these three bands. Since most grain comes to the river from the easternmost band, this weighting would yield an N value of 1.2. Col. Mudd directed that this value be applied to all agricultural products demands for barge transportation for all origin and destination pairs.
Colonel Mudd is not an economist and no economist involved in the study was involved in generating this N of 1.2 or has ever endorsed this number.
There are many flaws with this rationale, which I set forth in a telephone conference call with Col. Mudd, Rich Manguno, Gary Loss and others on June 22, 1999 and subsequent memorandum (Appendix 35). The most basic flaw is that this rationale is, among other things, based on an obvious mathematical error. The N value appears in the functions in the ESSENCE model as an exponent. Taking the linearly weighted average of exponents to three (or any number of) functions (the functions for three separate grain regions) does not yield the correctly weighted average of these combined functions.
On June 10, 1999, I sent a memorandum by email to Colonel Mudd, Gary Loss, Richard Manguno and Diane Karnish noting that General Anderson’s September 1998 directive extending the economics panel for the duration of the study had made the economics panel "responsible for" all economic products but the panel had not met for more than six months (Appendix 35). I asked Colonel Mudd whether he intended to call a meeting prior to public meetings on the navigation study scheduled for July 1999.
In response to this email, the first economic panel meeting since September 1998 was held via a teleconference on June 22, 1999 with all available panel members. Not all panel members were available on that date and I recall the participants were Col. Mudd, Rich Manguno, Wes Walker, Gary Loss, and I. On this call, Mr. Manguno briefed the panel members on the changes made to the N values and the theory behind the changes. During this call, Gary Loss also described the changes made to the major rehabilitation analysis and project cost estimates.
During the call, I explained to Colonel Mudd several reasons why his methodology for arriving at an N value of 1.2 was flawed, including most of the reasons set forth later by me in a memorandum dated June 29, 1999 (Appendix 35). Wes Walker expressed concern regarding all the changes that had taken place since the panel last met and requested more detail and time to examine the changes. The economics panel also questioned the use of short run rail elasticities in the model for goods other than grain as a substitute for long run water elasticities.
The panel requested details and explanations for why the already reviewed cost estimates for project construction had decreased and for why rehabilitation costs avoided had increased so dramatically over the levels the panel had been previously briefed. Col. Mudd requested written comments and recommended solutions be immediately forwarded to him and the project manager Gary Loss.
On June 23, Gary Loss provided panel members with a file setting forth the changes to the major rehabilitation numbers.
On June 24, I sent Gary Loss a memorandum, with copies to other economists and Colonel Mudd, requesting further explanation of the new numbers. On the same day, Gary Loss forwarded to me an email response from Roger Less, an engineer in the Rock Island District office, that provided only the most cursory explanation of the changes and that made no reference to any documented technical analysis. Copies of this correspondence are attached as Appendix 36. Wes Walker and I were the only panel members to send written comments. I sent my comments in a memo dated June 29, 1999. Wes Walker sent his comments on also on June 29, 1999. Our comments raised a number of objections to the changes and are attached as Appendix 35.
Three days later, on July 2, 1999, Colonel Mudd emailed Gen. Anderson recommending disbanding the economics panel for the UMR-IW system study. In an email dated July 4, 1999, Gen. Anderson so ordered. These emails are attached as Appendix 37.
In a June 1999 newsletter, the Army Corps presented its new preliminary benefit/cost analysis of the various project alternatives in a series of public meetings. This analysis for the first time now indicated that a project involving the immediate extension of seven locks now provided benefits greater than their costs. Copies of this presentation, which is publicly available, are attached as Appendix 38.

D. Developments Since the Summer of 1999

In July 1999 and through the summer, the Army Corps presented its new preliminary benefit and cost analysis of the various project alternatives in a series of public meetings.
There were relatively minor changes to the inputs and therefore economic analysis of project alternatives over the summer and early fall of 1999, which modestly changed the benefit/cost analysis of the various alternatives. At a November 18, 1999 meeting of the Governor’s Liaison Committee, Rich Manguno presented the analysis of benefits and costs at that time. This presentation shows that an analysis of extending five locks had a net positive economic benefit of $8.686 million per year, but of this $7.122 million were rehabilitation cost savings.
Despite all the changes directed by top Army Corps of Engineers officials, this analysis also showed that the alternatives that had the highest net positive benefits still involved postponing the start of construction of lock extensions until 2011. This is because, even under the Faucett traffic projections, the modest traffic congestion in the early years does not warrant the cost of lock extensions at that time and so drags down the overall benefit/cost analysis. A copy of Mr. Manguno’s "overheads" presented to this meeting is attached as Appendix 2.
The November 1999 GLC meeting was also the first time at which the study team provided an estimate of system wide environmental costs. The study team had early decided to separate the environmental analysis of the direct impact of extending locks, which could immediately displace some wildlife habitat, with the so-called system wide impacts of the overall project. These system wide impacts involved such impacts as mortality from increased barge traffic of fish. Their costs had to be added to the financial costs of any project. None of the benefit-cost analyses described so far, including the presentation made by Mr. Manguno at the November, 1999 GLC meeting, incorporated these system wide environmental costs.
On January 12, 2000, Dave Tipple sent out by email a slightly revised optimization study provided him two days earlier by Rich Manguno which still indicated that the alternatives with the highest economic benefits all involved postponing the start of lock construction until at least 2011. This analysis still does not account for system wide environmental costs, which would further postpone the time at which lock extensions were justified. And during this period, of course, traffic forecast projections could be verified. (Copies of this analysis are attached as Appendix 39.)
None of the benefit cost analyses described in this affidavit incorporated most environmental costs. The study team had early decided to separate the environmental analysis of the site-specific impact of extending locks, which could immediately displace some wildlife habitat, with the so-called system wide impacts of the overall project. These system wide impacts involved such effects as fish mortality from increased barge traffic. These system costs, which have been estimated to be far greater than the site-specific costs, must be added to the financial costs of any project, thus reducing the net economic benefits.
At a January Navigation Environmental Coordinating Committee meeting, Ken Barr presented what was then described as the preliminary estimate of system wide environmental costs for alternative J. This alternative involved the extension of 5 Mississippi River locks and the two Illinois River locks. This estimate put system environmental costs at $10.5 million per year. I have seen a copy of the handout listing this figure, and it is attached as Appendix 40). Mr. Manguno’s benefit and cost analyses set forth at the GLC meeting in November (Appendix 2) and identically in Appendix 39 indicate that the total net economic benefits for this alternative, not counting these environmental costs, is $11.026 million. In other words, at least as of January, 2000 the total average annual benefits for extension of seven locks was estimated to be $526,000.

V. SUMMARY

This section of my affidavit summarizes the above discussion.
The events show that top Army Corps of Engineer officials, including Major General Fuhrman, Major General Anderson, and Colonel Mudd, repeatedly rejected the advice of a broad range of Corps economists, and continually attempted to transfer management, production and review of economic products in an effort to find economists who could or would justify large-scale improvements. The events show that the model developed by the study team was first presented to a broad range of Army Corps of Engineers economists in October of 1997, who responded enthusiastically. It was then subjected to an Independent Technical Review, which had resulted by June 1998 in endorsement of the model while raising a few areas for added work. Despite this review, because of concerns that I was "out to shutdown the Corps," General Anderson established a new panel to completely review again the economic products and to oversee the future production of economic products. By August 1998, this new economic panel had endorsed the study and recommended an N value of 1.5 for grain (the key N value). Also at the direction of the study manager, a second model was used to verify the UMR-IW economic model under the direction of Dr. Mike Bronzini, which yielded similar results. In September 1998, the panel was essentially shelved and management given to Mr. Manguno under instructions to produce an analysis to justify large-scale improvements. Despite this instruction, Mr. Manguno continued to insist that he could not justify using an N value of less than 1.5 for grain in the model. He was directed to use an N value of 1.2 for grain. Then, when the economics panel reengaged, because it had not yet been formally abolished previously, and expressed disagreement with this direction, it was abolished. In short, all Army Corps economists involved in the study ultimately supported the ultimate analytic method, and none supported use of an N value less than 1.5 for grain.
The evidence also shows that public presentations or release of the preliminary economic analysis were scheduled as early as April 1998 and were repeatedly postponed when the economic analysis at that time showed that large-scale measures were not justified. Even in November 1998, when results were released, only a range of alternatives was presented. Moreover, such a range should presumably have encompassed a range around the most likely alternatives. Because the economic panel had recommended the use of an N for grain of 1.5, N values of 2.0 should presumably have been shown. But the highest N value represented was the 1.5. Again, when results were to be provided to the Economic Coordinating Committee and to the GLC in May of 1999, the presentations were postponed. Only after the analysis had been altered in May and June of 1999, were preliminary results released to the public.
In addition to critical changes to the N value, the events show deliberate efforts to distort the analysis of rehabilitation cost savings. An original engineering analysis reported no saved rehabilitation cost savings from expanded locks and this was reflected in the engineering handoff to economics in March 1988, which followed an Independent Technical Review. This conclusion remained in place despite frequent discussion of the topic by study managers until May of 1999. The evidence shows that the idea for changing this analysis was then introduced at a meeting with industry in May of 1999 by Ron Keeney, the Chief of Programs and Project management in the Huntington District, who acknowledged that the analysis to date showed no savings but who noted that his Division frequently found such savings helped to justify new projects. New information describing such savings was then provided to Mr. Manguno within days. No detailed documentation of the new analysis has ever been provided to the study team leaders or to the public and it has undergone no independent technical review. The public explanation of this change contradicts information provided to Mr. Marmorstein and me that the cost estimates of lock extensions included no rehabilitation costs. In contrast to the extensive engineering and economic analyses that indicated that no major rehabilitation for major components would be required until at least 2033, this recent public explanation also stated that rehabilitation would be required on many locks in 2015 based on the definition of rehabilitation as something that occurs every 25 years. In the economic analysis presented at the GLC in November 1999, the rehabilitation cost savings represented roughly 7/8ths of the net benefits of extending locks.
The evidence also shows that in the original economic analysis, industry would and could use self-help much of the time to reduce the congestion costs even without any new construction. The evidence shows that Mr. Manguno was instructed in January 1999 to restrict the model to allow self-help only 5% of the time. No documentation of any analysis has ever been made to support this arbitrary restriction.
The evidence also shows that the original costs for the expanded locks included a 35% contingency cost because of the innovative nature of the construction techniques. The Independent Technical Review of the work products of the engineering work group required that the contingency cost estimate be set at this level because of the untried new construction techniques and engineering plan. The evidence shows that this contingency cost estimate was arbitrarily lowered without any documented or independently reviewed additional analysis within sixteen days following a meeting with navigation industry representatives on May 12, 1999.
The evidence also shows that despite all these changes, and despite the use of projected increases in traffic demand that real world experience has shown to be greatly overestimated so far, the total net economic benefits are vanishingly small. For an alternative that involves immediate extension of seven locks, they amount to only around $500,000 per year. Other alternatives may have slightly higher or lower numbers. This means that if any of the directed changes had not been made (changes to the elasticity of demand for barging, i.e., the N values, changes to the rehabilitation cost savings, changes to the self-help analysis, changes to the contingency costs), expansion of locks would not have a positive benefit/cost analysis for the foreseeable future.
Finally, the evidence shows that even with all these changes, the alternative that would maximize net economic benefits would still be to postpone the start of lock expansion until at least 2011. That alternative would also allow experience to verify if traffic is in fact growing sufficient to justify lock extension at that time.