Evaluating the Performance of Environmental Streamlining: Phase II

table OF CONTENTS

1.0INTRODUCTION
1.1	Background and Purpose
1.2	Findings of the Phase I Research
1.3	Purpose and Reorientation of the Phase II Research
1.4	Organization of Report

2.0RESEARCH APPROACH
2.1	Research Approach Overview
2.2	Description of Data Sources
2.2.1	Northwestern University Transportation Library
2.2.2	Lexis-Nexis Database
2.3	Limitations of Available Data
2.3.1	Unavailability of Record of Decision Dates
2.3.2	Elimination of Select EISs from Database
2.4	Data Collection Methodology
2.4.1	Review of EIS Project Details
2.4.2	Development and Verification of Database
2.5	Statistical Analysis Methodology
2.5.1	Identification of Descriptive Statistics
2.5.2	Selection of Data Analysis Methods
2.5.3	Statistical Relationship Test Between Variables

3.0RESULTS OF RESEARCH
3.1	Descriptive Statistics on the Length of the NEPA Process
3.1.1	Overall Descriptive Statistics
3.1.2	Descriptive Statistics by Region
3.2	Descriptive Statistics on Other Factors Considered
3.3	Statistical Relationship of the Length of the NEPA Process with Other Factors
3.4	Use of Results as a Baseline for Evaluating Future Environmental Streamlining Initiatives

4.0SUMMARY AND CONCLUSIONS
4.1	Summary of Findings
4.2	Conclusions

5.0LIST OF PREPARERS

LIST OF tableS
table 1:	Final Database Variables and their Definitions
table 2:	Statistical Tests Conducted for the Length of the NEPA Process
table 3:	Length of NEPA Process (In Years)
table 4:	Descriptive Statistics of NEPATIME by Region
table 5:	Relationship of Geographic Categories With NEPATIME / SQRTNTIM
table 6:	Descriptive Statistics for Selected Variables
table 7:	Relationships Between Length of NEPA Process and Selected Other Regulatory Requirements
table 8:	Issues Variables and the Length of the NEPA Process

Appendices A through K are not available for viewing online at this time. Please contact Kreig Larson at FHWA kreig.larson@fhwa.dot.gov if you wish to receive them on CD-ROM.

Back to Top

EXECUTIVE SUMMARY

The National Environmental Policy Act (NEPA) directs federal agencies, when planning projects or issuing permits, to conduct environmental reviews to consider the potential impacts on the environment by their proposed actions. The NEPA umbrella includes the use of federal funds for construction of highway and bridge projects falling under the purview of the U.S. Department of Transportation, Federal Highway Administration (USDOT / FHWA).

NEPA significantly contributes to the environmental quality of federally funded projects. However, because the NEPA process requires coordination and consultation with other agencies as well as the public in making decisions on projects, it is widely perceived as complex and time-consuming, and thus as a major source of project delay and cost increase. This perception is further magnified when a project is subject to the requirements of other laws and regulations that are to be addressed as part of the NEPA project development process. Such regulatory complexities have led to a national initiative to streamline the environmental process and expedite project delivery. The Transportation Equity Act for the 21st Century (TEA-21) and, more recently, Executive Order 13274: Environmental Stewardship and Transportation Infrastructure Project Reviews emphasize making transportation decision-making and environmental review processes more effective and efficient.

To evaluate the relative success of ongoing and future environmental streamlining efforts pursuant to TEA-21 and other mandates, FHWA and the Louis Berger Group first undertook extensive research to establish a historical baseline related to the NEPA time frame. The initial phase of that research was published in January 2001. Phase I met most of its goals, except for determining the cost aspects of both the processes for NEPA and for overall project delivery; this was attributed to limitations of the data and to the difficulty of addressing cost compatibility over the 30-year research period. It was clear that a more in-depth and targeted analysis of compliance time frame trends, as well as the factors that impact time frame requirements, was needed.

The purpose of this Phase II study was to ascertain if the baseline results found in Phase I would be repeated, or if the new analysis produced a baseline that better reflected those factors that impact the NEPA process for transportation projects.

Conclusions

The Phase II study achieved the basic goals of expanding the baseline condition and identifying continuing trends. For example, the study shows that in comparison to earlier decades, the amount of time to complete the NEPA process has continued to increase in recent years. Findings further suggest that substantial opportunity exists for improvement in the overall process, but that improvement may only result in a total NEPA process time representative of an earlier decade.

However, this study was not able to achieve the identification and confirmation of factors and conditions that may have a direct or indirect impact on the NEPA process. Although the process seems to vary by broad geographic region, it does not seem to vary in relation to the majority of other variables tested in this data set. Some factors were found to have a statistically significant relationship to the time needed to complete the NEPA process, but the relationship was either weak or difficult to explain in a practical sense. Other factors were not found to have a significant statistical relationship to the time needed to complete the process, although they actually appear to influence it. The factors identified in this study as being related to the time needed to complete NEPA were different from those factors identified during the Phase I study; the reason for this is not entirely evident. In fact, the only factor that was found to have a noticeable effect on NEPA process time in both phases of the Study was the presence or absence of a Section 4(f) evaluation. Even so, in the current study this factor was not found to have a statistically significant relationship with NEPA process time.

Given the lack of statistical variation in the length of the NEPA process when considered in relation to other project- and process-related factors, it can perhaps be concluded that the process may have become more strongly affected by external social, economic, and attitudinal factors associated with broad geographic regions of the country. If true, this theory would account for the inability to explain with any certainty the causal relationships of specific factors tested in this study. Although such a theory may make it difficult to take better control of the process by refining or varying specific project- or process-related factors, it would make clear that environmental streamlining efforts should place more emphasis on social and economic conditions that could affect the process in a given location. If nothing else, it reinforces the overall complexity of the NEPA process and the broad zone of influence that external conditions may have on its application.

Back to Top

1.0 INTRODUCTION

1.1 Background and Purpose

The National Environmental Policy Act of 1969 (NEPA) marked the beginning of the environmental review process for all federal actions, including the use of federal funds for construction of highway and bridge projects falling under the purview of the U.S. Department of Transportation, Federal Highway Administration (USDOT / FHWA). It is the goal of NEPA to ensure that all federal agencies take the potential environmental implications of their actions into account in advance of performing those actions, and to ensure that all reasonable means are taken to minimize the adverse effects resulting from those actions. In the case of highway and bridge projects specifically, NEPA ensures that FHWA will make informed decisions with respect to the environment when considering the need for, and proposed alignment and design of such projects.

Although NEPA has wide ranging benefits when it comes to the proper delivery of federally-funded projects, it also has a long history of being criticized due to its perception as an overwhelming, time-consuming process. In this regard, NEPA has long been blamed as a major source of delay and inflationary cost increase, especially when the entire project delivery process from beginning of the NEPA process to the completion of construction is taken into consideration. This criticism has been especially true when the most complex and time-consuming level of environmental documentation required under NEPA, the Environmental Impact Statement (EIS), is involved. The problem is further magnified when other laws and regulations are required and addressed under the overall NEPA umbrella. Special requirements such as Section 4(f) evaluation, Section 106 review, Section 7 biological assessment, Section 404 wetlands permitting, Section 9 bridge permitting, etc. further add to the overall perception of NEPA as a major source of project delay.

In recognition of the widespread perception of project delays, regulatory complexities and the potential for improvement in implementing the NEPA process, a national initiative for introducing "environmental streamlining" into the process was included in the Transportation Equity Act for the 21st Century (TEA-21, P.L. 105-178) in 1998. Section 1309 of that Act directs DOT to "develop and implement a coordinated environmental review process for highway construction projects." This requirement is especially applicable to projects for which an EIS is to be prepared pursuant to NEPA.

More recently, in September 2002, President Bush signed an Executive Order that requires Federal agencies to take appropriate actions to promote environmental stewardship in the Nation's transportation system and expedite environmental reviews of high-priority transportation infrastructure projects. The Executive Order also creates a new "Transportation Infrastructure Streamlining Task Force" to more closely coordinate federal reviews on projects while simultaneously stressing the importance of improved environmental stewardship at all levels of government. Similarly, both bodies of Congress have recently introduced bills designed to streamline the environmental process and expedite the project delivery process.

In order to evaluate the relative success of ongoing and future environmental streamlining efforts pursuant to TEA-21 and other mandates, it was first deemed necessary to establish a historical baseline against which to compare NEPA time frames resulting from such efforts, thereby measuring continuous performance. However, until recently, limited information has been available regarding the actual time frame that has been required for preparing NEPA EIS documents from beginning to end, and virtually no information existed to place that time frame into the context of the overall project delivery process. Given the limited information available in these regards, FHWA undertook an extensive research endeavor to establish a historical baseline related to NEPA time frame, specifically NEPA EIS time frame. The first phase of that research was published in January 2001 while the second phase is the subject of this report.

1.2 Findings of the Phase I Research

The Phase I research was entitled Evaluating the Performance of Environmental Streamlining: Development of a NEPA Baseline for Measuring Continuous Performance, prepared by The Louis Berger Group, Inc. for FHWA. The purpose of that initial effort, from the outset, had been to provide a better understanding of the impacts of the NEPA process on the total time and cost involved in delivering a federal-aid highway or bridge project. Several basic goals for the initial research phase were identified as follows:

• To establish a baseline against which to assess future environmental streamlining efforts;
• To identify what component of the time and cost of the overall project delivery process is attributed to NEPA;
• To identify and assess a wide variety of factors and conditions that may have a direct or indirect impact upon the NEPA process and the project delivery process as a whole, and to determine the predictive strength of the relationship between NEPA and those factors; and
• To identify appropriate data sources and formats that could be utilized for developing the baseline conditions, as well as for monitoring future progress of environmental streamlining.

That research met most of the identified goals, except for the cost aspects of both NEPA and the overall project delivery process. Due to limitations of the data and the difficulties of dealing with cost compatibility over a 30-year research period, the portion of the overall project delivery process cost that is attributed to NEPA could not be ascertained. In addition, the predictive ability of the statistical correlations found to exist between NEPA and a variety of other factors is considered to be relatively limited, although several factors appear to have had some historical influence on the NEPA time frame.

An extensive data collection process was developed and executed as part of the Phase I research, and a statistical analysis was completed using the database. A summary of the Phase I results is provided below:

• Based on the first 30 years of NEPA compliance, the typical length of time for preparing an EIS pursuant to NEPA has been either 3.0 or 3.6 years, depending on whether the median or mean values are used, respectively;
• The length of time for preparing an EIS pursuant to NEPA has varied between former FHWA regions, with the greatest time required in Region 1 (4.5 or 4.7 years, based on median or mean values, respectively) and the least time required in Region 10 (1.0 or 2.2 years, based on median or mean values, respectively);
• For those projects in which an EIS pursuant to NEPA was required, the mean time required for the entire project development process has been approximately 13.1 years;
• For those projects in which an EIS pursuant to NEPA was required, the NEPA process accounted for approximately 27 to 28% of the total time required for the entire project development process, depending on whether the median or mean values are used, respectively;
• The length of time for preparing an EIS pursuant to NEPA has varied between each of the three decades that have occurred since NEPA was implemented, ranging from a mean of 2.2 years in the 1970s to a mean of 5.0 years in the early 1990s;
• The length of time for preparing an EIS pursuant to NEPA has varied depending on whether or not a Section 404 permit was also required, ranging from a mean of 2.4 years when no Section 404 permit was required to a mean of 4.3 years when a Section 404 permit was required;
• The length of time for preparing an EIS pursuant to NEPA has varied depending on whether or not a Section 4(f) approval was also required, ranging from a mean of 2.8 years when no Section 4(f) approval was required to a mean of 4.7 years when Section 4(f) approval was required;
• The length of time for preparing an EIS pursuant to NEPA has varied depending on the number of agency meetings held, ranging from a mean of 2.4 years when fewer than three agency meetings were held to a mean of 4.5 years when three or more agency meetings were held; and
• The length of time for preparing an EIS pursuant to NEPA has varied depending on whether or not noise has been an issue, ranging from a mean of 3.2 years when noise was not an issue to a mean of 4.4 years when noise was an issue.

1.3 Purpose and Reorientation of the Phase II Research

The desire to uncover more information that could further build on the conclusions of the initial study led FHWA to pursue a second phase of research. The second phase, which is the subject of this report, has changed its focus somewhat from that of the first phase, since most of the original goals of that research had already been met. Given that a major aspect of the first research phase was to assess the relationship of NEPA and the overall project development process, and that the basic relationship had already been established, it was not considered to be prudent to continue focusing on that relationship in this follow-on study. Similarly, since the entire 30-year period that NEPA has been in effect was the focus of the original study, it was determined to be unnecessary to cover the same period with an expanded sample size.

The basic intent of developing and refining a NEPA baseline time frame, however, has not changed in this second phase. It is also still the intent of this research to provide a baseline against which future EIS projects can be compared in order to measure continuous performance of environmental streamlining efforts. What has changed is the breadth of the research in terms of historical time period considered and the level of detail sought to refine the baseline for the universe of projects within that revised time period. As a result of these fundamental changes, the methodology used in this Phase II study has also changed from that used during Phase I. The main differences between the two phases are listed below:

• Phase II has a revised focus on only the more recent EIS projects (i.e., those completed from 1995-2001) instead of the entire 30-year period following inception of NEPA as in Phase I;
• Phase II involves analysis of the total universe of EIS projects completed during the limited period of study instead of a random sample representing the total universe of EIS projects during the broader period of study as in Phase I;
• Phase II focuses only on the NEPA process itself, instead of the relationship of NEPA to the overall project delivery process as was considered in Phase I;
• Phase II eliminates the Phase I requirement that construction of the EIS projects must have been completed in order to consider them, since the overall project delivery process is not to be considered;
• Phase II improves upon the identification of a formal starting point for the NEPA process for each studied project (i.e., Notice of Intent date), thereby reducing the level of subjectivity inherent in Phase I; and
• Phase II eliminates the need to address those factors found during Phase I to be of little value in explaining or describing the effect on the NEPA compliance time frame.

By incorporating the above changes into the basic research structure of Phase II, it was anticipated that a more concise and in-depth view of the NEPA compliance time frame trends and the factors that impact upon NEPA time frame requirements could be performed. Given this basic assumption, the overall contribution of this second research phase to NEPA baseline development and refinement will have substantial merit.

1.4 Organization of Report

Following this Introduction, Section 2.0 provides a detailed discussion of the Research Approach used for this study. Specific aspects of the Research Approach addressed include: a Research Approach Overview; Description of Data Sources; Limitations of Available Data; Data Collection Methodology; and Statistical Analysis Methodology.

Section 3.0, Results of Research, provides the basic findings of the study. These findings include: Descriptive Statistics on the Length of the NEPA Process; Descriptive Statistics on Other Factors Considered; Statistical Relationship of Length of the NEPA Process with Other Factors; and Use of Results as a Baseline for Evaluating Future Environmental Streamlining Initiatives.

Section 4.0 provides a discussion of Conclusions and Recommendations. This section presents a summary of the findings and offers recommendations, as appropriate.

Section 5.0 provides a List of Preparers.

A series of Appendices that assist in illustrating details of the research survey universe development, data collection, and statistical analysis of the data conclude the report. These Appendices are denoted as Appendix A through Appendix K and are available from FHWA on a separate CD-ROM.

Back to Top

2.0 RESEARCH APPROACH

2.1 Research Approach Overview

This current research study, like the original Phase I study, has been designed to provide a comprehensive, objective, and statistically-based approach to refining a baseline regarding the length of time required to comply with NEPA. Similar to Phase I, this Phase II research only deals with projects subjected to an EIS level of documentation, rather than those involving Environmental Assessments (EAs) or Categorical Exclusions (CEs).

As stated in Section 1.3 above, there are a number of changes that have been incorporated into the overall approach used for the Phase II research in comparison to that used in Phase I. These changes are discussed in further detail below.

• Concentration on More Recent Projects Only — Whereas the Phase I research was very broad in terms of its historical perspective, covering the entire period since inception of NEPA in 1969, this second phase concentrates only on projects in which the EIS document was completed between the years 1995 and 2001. Inclusion of these more recent projects in Phase II results in a totally different data set from that used in Phase I, since no projects having an EIS document completed later than the early 1990s was addressed in that study. By concentrating only on the more recent projects that were completed and approved through the NEPA process, it is perhaps possible to draw comparisons between these projects and the earlier NEPA projects studied in Phase I.
• Analysis of the Total Universe of NEPA Projects— Whereas the Phase I research was based on a random, stratified sample of EIS projects completed during the study period, the Phase II research is based on the entire universe of EIS projects completed and approved during the study period. Given the smaller study period involved in this current effort in comparison to the earlier effort, the ability to consider the entire universe is much more practical and feasible. Therefore, the entire list of 250 EIS projects that were identified as having been completed during the 1995 — 2001 period have been included in Phase II. By including the entire universe of projects, potential statistical issues related to sample size and error do not exist. As a result, the statistical findings that result from the research should theoretically reflect the real condition, rather than a sample condition that could be argued doesn´t match with the real world.
• Concentration on NEPA Process Only— Whereas the Phase I research took the overall project delivery process into account, the Phase II research does not consider the overall project delivery process at all. Instead, this current research focuses only on the actual NEPA process and other regulatory requirements that fall under the NEPA umbrella, which together, only comprises a portion of the overall project delivery process. Since the relationship between the NEPA time frame and the overall project delivery process time frame was already established in the earlier study, it was concluded that it was unnecessary to consider that relationship again in this study. As a result of this fundamental change, it became unnecessary to collect and integrate specific information about the overall process into the database. As a further result of this change, the need to deal with FHWA´s Fiscal Management Information System (FMIS) database, which was an integral part of the Phase I research, was also eliminated. This change also eliminated the need to focus only on those EIS projects that have been fully constructed.
• Elimination of Project Construction Completion Requirement — Whereas the nature and intent of the Phase I research required that it focus only on EIS projects that have already been fully constructed, that requirement did not apply to Phase II. In fact, the status of each project's design and construction was unnecessary to know for the current research. It was the elimination of the construction completion requirement in Phase II that allowed the research to re-focus on the more recent NEPA projects, since it is very unlikely that many of those projects have actually been constructed yet.
• Refinement of NEPA Starting Date — Whereas the NEPA starting date used in defining the total time frame for each EIS project in Phase I was based on information of variable quality as presented in the actual EIS documents, Phase II uses the official Notice of Intent (NOI) date. The ability to standardize each project's start date in this current research eliminates the subjectivity that was inherent with the former approach. The NOI dates could not previously be considered because access to the primary source of data for such information was not available at the time.
• Elimination of Extraneous Variables— Whereas the Phase I study included a total of 74 separate data variables in the database, those that were found to be of little value in explaining or describing the effect on the NEPA compliance time frame were eliminated from consideration in Phase II. For instance, this current research only addresses time aspects of NEPA, rather than cost, as it was demonstrated during Phase I that cost really could not be considered in a cost-effective manner given severe limitations and general unavailability of cost data and difficulties in comparative cost analysis over a period of time. Other variables for which it had previously been shown to be difficult to obtain meaningful data have also been eliminated.

As a result of the above changes, the overall methodology used in Phase II is generally much less complex than that used in the Phase I research. This is especially true in light of the elimination of the need to use the FMIS database system, which was difficult and time-consuming in terms of integration of those appropriation records into each EIS project. Other changes that reduced the complexity of this effort was the use of the entire universe of projects during the study period rather than developing a random and statistically-valid sample, as well as the elimination of the need to determine which projects had already been fully constructed. However, the reduction in complexity was countered by the substantial increase in total number of EIS projects (from 100 to 250) for which detailed data had to be collected and compiled. In addition, clarification and standardization of the official NEPA start date for each project resulted in increased effort as well.

The basic steps of the overall methodology used in this Phase II study, in the order that they were followed, are presented below.

• From the list of EIS projects available from the Northwestern University Transportation Library, the total universe of EISs published between the years 1995 and 2001 was identified;

• Obtain and review the EIS document(s) associated with each potential project to be included in the database in order to identify locational and component parameters of the project;
• Identify EIS start dates for each project using NOI dates available from the Federal Register;
• Enter all relevant project data collected from the EIS document(s) and the Federal Register into a computerized database, and provide visual quality control checks for anomalies in the entered data, and revise as necessary;

• Perform descriptive statistics analysis, including maximum and minimum values and, in some cases, frequencies, to identify additional anomalies in the data and to establish descriptive parameters of each data variable;
• Perform exploratory data techniques to look for differences in descriptive parameters among different former FHWA regions;
• Create new data variables as necessary to complete statistical analysis by manipulating entered data variables;
• Determine if the data appear to be normally distributed, which would allow statistical analysis to be appropriately performed or, if not, transform data to obtain surrogates that are normally distributed and can be appropriately analyzed;
• Test for differences among former FHWA regions with regard to length of the NEPA process and other variables;
• Examine relationships, including correlations between NEPA process time and other data variables; and

• Utilizing the results of the statistical analysis, identify a refined NEPA process baseline or set of baselines under various conditions against which to evaluate future environmental streamlining efforts.

Tracing through the individual steps presented above provides a general indication of the overall methodology utilized in undertaking this Phase II research study. However, each individual component of the process involved its own set of substeps and analyses. A detailed discussion of each of these steps, including specific variables used, data limitations and problems encountered, is provided in subsequent subsections of this report.

2.2 Description of Data Sources

There were two main sources of data used for undertaking this Phase II research: 1) the Northwestern University Transportation Library; and 2) the Lexis-Nexis on-line commercial database service. The composition, format, and availability of these data sources are discussed in detail below.

2.2.1 Northwestern University Transportation Library

As was the case with the Phase I study, the Northwestern University (NWU) Transportation Library and its World Wide Web site served as the starting point for undertaking this research. The NWU Library contains more than 4,000 EIS documents prepared for Federal-aid highway projects, dating back to 1970 when the requirements for NEPA documentation first went into effect. For the Phase II study, only those EISs completed during the period beginning in 1995 and ending in 2001 were requested, a total of seven years of EIS documents. It was found that the NWU Library contains a total of 250 Final EIS documents related to FHWA projects from that period.

Recognizing that the NWU Library collection of FHWA EIS documents could in fact be somewhat different from the total number of FHWA EISs actually completed during the study period, an analysis was performed to verify the relative completeness of its collection. Given that the Phase II research is based on the entire universe of EISs during the 1995 — 2001 period, it was decided that a comparative review of the NWU Library documents should be made against those EISs identified by the U.S. Environmental Protection Agency (EPA) in its list of Notices of Availability and by FHWA in its annual surveys of completed EISs by state since 1998. In this manner, an assessment of the degree of completeness of the NWU source could be verified.

In the case of the comparative review between the NWU source and the EPA source, an assessment for each year between 1995 and 2001 was conducted. It was found that in addition to the 250 project document sets available at the NWU Library, 33 EIS projects were identified from the EPA website which were not identified by the NWU Library. On the other hand, 27 of the EIS projects identified by the NWU Library were not identified on the EPA website.

In the case of the comparative review between the NWU source and the FHWA annual breakdown of EISs between 1998 and 2000, it was found that 42 documents listed by NWU as having been completed during those three years are missing from the FHWA database. Alternately, a total of 15 documents on the FHWA list were found not to be included on the NWU list of documents, at least during those same years.

It should be noted that many of the documents that exist on both the NWU and the FHWA lists do not match up in terms of completion year. In all such cases, the FHWA list identifies a later date than the NWU list. For instance, a particular project EIS identified as being completed in 1998 on the NWU list may be identified as being completed in 1999 on the FHWA list. This difference is probably attributable to the fact that the NWU list reflects the date of Final EIS signature while the FHWA list reflects the date of the Record of Decision. In a few cases, the EIS completion years shown on the FHWA list were two or more years after the completion years indicated on the NWU list.

The staff from the NWU Library was unable to provide any conclusive reason for the absence of some documents within its collection. In the case of those documents identified on the EPA list which were missing from the NWU list, it is suspected that the EPA, who provides the EIS documents to the NWU Library for cataloging, had merely run out of copies before they could be forwarded. Although it was concluded that some EIS documents that were completed during the 1995 — 2001 period are missing from the NWU Library and its computerized database, its list of 250 EIS projects appears to be fairly representative of the total actual universe of such projects during that period. The differences between the NWU Library list and both the EPA and FHWA lists tend to cancel each other out, indicating that the NWU Library documents collection is adequate for representing the universe of FHWA EISs. It should be noted that no discernible pattern of difference was identified between the NWU database and either of the other two sources with regard to the geographic location or project type of the missing EIS projects. Having concluded the relative completeness of its database, the NWU Library again served as the backbone of this Phase II research, just as it had during Phase I.

At the outset of the Phase II study, contacts made and procedures used during the Phase I research were reestablished with the NWU Library. The NWU Library staff prepared and arranged the entire universe of EIS projects for the 1995 — 2001 period in reverse chronological order. Each entry on the list generally included: the title of the EIS project, which often included an indication of its location as well; date of Final EIS signature; EIS call number, which also identified the state responsible for implementing the EIS; the specific documents on file for each EIS project; and the Library's availability status for each document. A list of all 250 EISs available and requested from the NWU Library is provided in Appendix A.

Upon receipt of the list of 250 EISs from the NWU Library, each EIS project was assigned a number which was provided to the NWU Library in a manner that would facilitate the requesting and receiving of documents for review. A system was developed to ensure that an adequate number of documents were requested and received each week in order to keep the research on schedule. The request for EISs was generally made via e-mail. Upon arrival, the documents were reviewed, analyzed and sent back, usually within the same week that they were received. The EIS document sets available from the NWU Library generally included the Draft EIS, Final EIS, Supplemental Draft and/or Final (as appropriate) and, often, Technical Appendices.

Although the process of requesting and receiving EIS documents from the NWU Library generally worked well, some delays in shipping, incomplete shipments and missing volumes did occasionally occur, which resulted in delays during this phase of the process. The delays were always short, and all the proper documents were received.

2.2.2 Lexis-Nexis Database

One important type of data that was not uniformly available from the actual EIS documents was the NOI date for each project. As a result, a different data source was required to retrieve that data. NOI dates for most of the 250 EIS projects were found in the Lexis-Nexis database system, which is a commercial service that provides legal information, news, public records and business information, including tax and regulatory publications, in on-line format. NOI on-line retrieval was performed using the system database of Federal Register documents. In a number of cases, the EIS project title used in the Federal Register documents differed somewhat from that used in the actual EIS document; however, even in those cases, the two datasets were ultimately able to be matched.

Although NOI dates for all 250 projects were found within the Lexis-Nexis database, there were approximately twenty occasions where the NOI dates and/or other milestone dates included in the on-line Federal Register (e.g., the Draft and/or Final EIS signature dates) did not correlate with those identified from the actual EIS documents obtained from the NWU Library. For instance, the NOI date stated in the Federal Register may have post-dated the signature date of the actual Draft or Final EIS document, which obviously would not reflect a proper chronology. The most common reason for this discrepancy was when a Supplemental EIS (either Draft or Final) document was required for a project. It was often the case that a new NOI date was issued and published in the Federal Register for a supplemental document, although it was not immediately obvious that such was the case.

In order to ensure that these projects were considered for their full NEPA time, proper documentation of the NOI date was obtained. Two different methods were used for obtaining the NOI dates for these particular cases. In most cases, the original NOI dates were provided within the text of the actual EIS document, so that the Federal Register-listed date was not even used. However, in cases where the NOI date could not be found within the EIS document, inquiries were made with the appropriate State DOT. After expending, sometimes, a substantial effort to locate the appropriate individual with knowledge of that particular project, or at least with knowledge about how to access information on that project, the State DOT was generally able to provide the NOI dates regarding these remaining few projects. In five cases, however, neither of these approaches led to usable NOI dates (for more about how these projects were addressed in the research, see Section 2.3.2 below).

Upon retrieval of the NOI dates from the Federal Register documents or, when necessary, the EIS documents or relevant State DOTs, they were saved into a central file for reference purposes. A sample of a Federal Register page from the Lexis-Nexis system with NOI information included is provided in Appendix B.

2.3 Limitations of Available Data

As indicated in Section 2.2 above, the availability of the two primary data sources used for this research (i.e., the Northwestern University Transportation Library and the Lexis-Nexis database system), was critical to the ability to even undertake this research from the outset. However, even though these two data sources allowed the research to move forward, there were some inherent limitations, as discussed below.

2.3.1 Unavailability of Record of Decision Dates

One type of data that was found to be unavailable from any central source was the Record of Decision (ROD) date, which is the official completion date of the NEPA process. During the initial Phase I study, inquiries were made with FHWA headquarters, the U.S. Environmental Protection Agency (USEPA), and the Council on Environmental Quality (CEQ) to determine if there is a central source that exists for identifying this date for each EIS project, and determined that such a source does not exist.

It was initially hoped that the on-line Federal Register database available through the Lexis-Nexis system could possibly include ROD information. However, it was verified during the preparation of this second phase of research that ROD dates are not generally recorded in the Federal Register, thereby eliminating the potential for using that source in this regard.

As was the case during the Phase I research, the possibility of uniformly contacting FHWA division offices to obtain ROD dates for each EIS project was considered but ultimately rejected based on the conclusion that the process would not have been cost-effective. This conclusion was based on several factors: the amount of time likely to be required to find the person(s) who has the information; institutional memory problems; and the possibility of receiving inconsistent data that could lead to statistical bias.

The potential for using the ROD dates available from FHWA´s surveyed list of EISs by state since 1998 was also explored. However, as discussed in Section 2.2.1, there were differences between the NWU database of EIS projects and the FHWA database of EIS projects, which made consistent use of the FHWA ROD dates impossible. Since the FHWA database only covered the period between 1998 and 2000 at the time of data collection for this research, it is likely that many of the 42 documents listed by NWU as having been completed during those three years did not have official ROD dates until at least 2001, the results of which were not available at that time. In addition, a total of 15 EIS projects shown on the FHWA list to have approved ROD dates during the 1998 to 2000 period did not appear on the NWU list of EIS projects, which served as the primary data source for this research. Due to these differences between the two data sources and the inability to collect consistent ROD date information, it was concluded that the use of ROD dates to define the end point for some projects and the use of Final EIS signature dates for other projects would be inappropriate.

It should be noted, however, that FHWA´s Office of Project Development and Environmental Review conducted a spot check of the ROD dates provided by the various FHWA Division offices for a total of 68 of the 244 projects included in this study. The project documents in question were completed (i.e., FEIS signed) during the years 1999 to 2001. The period between the approval of the FEIS and the approval of the subsequent ROD was found to range from 1 to 9 months, with the mean being approximately 3.3 months.

As a result of the various data limitations, it was decided, similar to the Phase I study, that the end points of each EIS project should be based directly on information available in the EIS documents themselves. In this manner, the end points were easy, given that every Final EIS document has an FHWA signature date identifying when it was completed. It was decided that the official date on the Final EIS was selected as the end point, rather than arbitrarily attempting to add some additional time period (probably a few months) after the signature date to reflect the ROD date. Although it is likely that most RODs were approved within a few months of the signature date, there is no basis to verify that assumption without conducting a statistical analysis to further test the assumption. Therefore, the EIS end date used for each of the projects included in the research universe was the date of the Final EIS signature. Both the month and year of the signature date were recorded for use in the analysis.

Therefore, the length of the NEPA process for purposes of this research was based on the period of time between the official NOI date and the signature date for the Final EIS. Although this time frame is slightly underestimated in comparison to the real time frame because it does not reflect that relatively short period after the Final EIS signature date and the ROD date, it is adequate for the purposes of this research and is consistent for all projects. This process, at least, is a refinement over the one used in the Phase I research, where access to the NOI date was also unavailable, thereby resulting in a somewhat subjective start date for each project as well.

2.3.2 Elimination of Select EISs from Database

There were a total of six projects out of the 250 projects included on the NWU Library database for the 1995 — 2001 period that, due to a variety of special circumstances, had to be eliminated from use in the Phase II study. Five of these projects, as discussed in Section 2.2.2, were eliminated due to inability to obtain accurate information regarding NOI date. For instance, EIS number 960438 (New Jersey) was eliminated from the database because no record of any NOI date for this project was able to be found in the on-line Federal Register records. It is unclear why such documentation could not be located.

Similarly, four other projects, including EIS numbers FHWA-WV-990063, FHWA-WV-960180, FHWA-IA-870460 and FHWA-MN-980115, were all eliminated due to the fact that no NOI date could be found reflecting the start of the original EIS process. The only NOI date that could be found for those projects was related to a follow-on Supplemental EIS process. Since an official NEPA start date could not be identified, it was concluded that those projects should not be considered in the research.

The final project, EIS number FHWA-CA-000205, was eliminated due to its total NEPA time of twenty years. This length of time was substantially longer than for any other project in the database, and did not reflect any type of Supplemental EIS process within that time frame. Technically then, this project was not eliminated because of limitations in the data, but because the data that were collected created a true anomaly in the compiled database. In order not to incorrectly skew the statistical analysis conducted on the database, it was determined that it was best to eliminate this project from further consideration. This is a typical and acceptable statistical procedure when anomalies in the data are identified, even if those particular data are correct and explainable.

As a result of the elimination of the above six projects from further consideration, a total of 244 projects were available for analysis.

2.4 Data Collection Methodology

2.4.1 Review of EIS Project Details

The EIS documents for each of the projects included in the database were reviewed, and the pertinent data were extracted and recorded using a project data sheet. Types of data recorded on the project data sheets included: project setting and description; milestone dates within the NEPA process; agencies involved in the NEPA process; permits required; environmental issues/controversies; mitigation measures proposed; and details of the public involvement process. A copy of the project data sheet form used in this regard is provided in Appendix C.

Specific types of information that were recorded on the project data form, and the criteria used to collect these data are described below.

• Reviewer — This field records the name of the person conducting the EIS review.
• Date — This field shows day, month, and year of data sheet completion.
• Project Name — The project name is the title as it appears on the Final Environmental Impact Statement.
• Location — Gives the location of the EIS project by city, county, and state.
• Generalized Land Use — Land use was determined by reading the Land Use section in the Affected Environment chapter and by interpreting the land use map. Three categories of land use were used:
  • Rural: Any project that has land use dominated by agriculture or vegetated areas.
  • Small City: Any project that lies within a city having low to medium population statistics or a suburban area.
  • Urban: Any project that lies within a major metropolitan center.

• Generalized Setting — The project´s status with respect to a Metropolitan Statistical Area (MSA) was identified. To determine if a project was within an MSA, the county or counties comprising the project area was/were compared and referenced to the nationwide MSA table. The project was either within an MSA, outside an MSA, or both.
• Population Change — Population change was determined in either of two ways. The first way was to find the population data for the area, and then find the percentage difference between the 1980 population statistics and the 1990 population statistics, dividing by ten to find annual rate of change. The second way, if raw data were not given, was by searching the Socioeconomic section of the Affected Environment chapter of the EIS for any information pertaining to population characteristics. If neither of these procedures was possible, it was rated "NA."
• Type of Improvement — Type of improvement as either a highway or bridge project, as well as specifics of that project were derived by reading and understanding the Preferred Alternative. Many times a description of the Preferred Alternative was found on the Final EIS Signature Page. If it was not there, it was usually explained in the Executive Summary. More clarification was found within the Preferred Alternative section of the Alternatives Considered chapter.
• Length — The length of a project was often given on the Final EIS Signature Page in conjunction with a description of the Preferred Alternative. If the length was not on the Final EIS Signature Page, then it was either in the Executive Summary or within the Preferred Alternative section of the Alternatives Considered chapter.
• Land Acquisition Required — The amount of land required for a project was not always stated. Many EISs would state acquisition numbers for large impacts such as wetlands or farmland, but would not say exactly what the total acreage taking would be. In some cases, however, the total amount of acquisition would be clearly stated. This information was usually found within the Executive Summary or the Land Use section of the Environmental Consequences chapter. For those projects where land acquisition was not stated, an "NA" was recorded.
• Notice of Intent — Notice of Intent (NOI) dates show exactly when a project began the official NEPA process. The NOI dates are generally published in the Federal Register. Searching for the NOI dates was accomplished as discussed in Section 2.2.2 above. Some EISs identified the NOI date in the Public Outreach section. Some even included a copy of the NOI announcement page as it appeared in the Federal Register.
• Other Date Indicating Start of NEPA Process — This date, which generally involved the date of initial scoping meetings, further indicated when the NEPA process began. However, when a NOI date was available, this field had little value.
• DEIS Signature Date — This was the date found on the DEIS Signature Page that indicated the exact day an FHWA representative signed the Draft EIS.
• FEIS Signature Date — This was the date found on the FEIS Signature Page that indicated the exact day an FHWA representative signed the Final EIS.
• Project Sponsor — The project sponsors were the primary agencies involved at the federal and state levels. This information was found on the Final EIS Signature Page with the federal agency listed above the state agency.
  • Federal Agency: The federal agency was always FHWA.
  • State Agency: The state agency was usually that state´s transportation agency.
• Cooperating Agencies — Cooperating Agencies were government agencies that were asked to play a major role with regards to the development of the EIS. A Cooperating Agency was generally included when a project caused specific and distinct impacts under its jurisdiction, or required other regulatory approvals by that agency. Cooperating Agencies were generally identified either on the Final EIS Signature Page or in the Final EIS Summary. In cases where they were not specifically identified in either of those two locations, they were sometimes identified in the Public Involvement section of the EIS.
• Commenting Agencies Number — Number of Commenting Agencies were government, businesses, and local groups that have written formal comment letters in regards to the information presented in the Draft EIS. These agencies were generally identified in a list of Commenting Agencies included in the Public Involvement section of the EIS. If such a list was not available, the Number of Commenting Agencies was then determined by counting the formal comment letters included in an appendix to the EIS document. The Commenting Agencies were categorized as follows:
  • Federal: Any agency that falls under federal jurisdiction.
  • State: Any agency that falls under state jurisdiction
  • Local Any agency that falls under local jurisdiction. Also included in this number were businesses, and local groups.
• Total Number of Cooperating/Commenting Agencies — This number was derived from the total number of cooperating agencies plus the total number of federal, state, and local commenting agencies. However, this entry is not part of the final database.
• USEPA Rating — The U.S. Environmental Protection Agency (EPA) has developed a rating system which summarizes its level of concern regarding a particular action, based on a given project´s Draft EIS document. The ratings are a combination of alphabetical categories that signify EPA‘s evaluation of the environmental impacts of the proposed project, and numerical categories that signify an evaluation of the adequacy of the EIS. There are four different rankings for the alphabetical portion of the rating:
  • LO (Lack of Objections), where the review has not identified any impacts that would necessitate changes to the Preferred Alternative.
  • EC (Environmental Concerns), where the review had identified environmental impacts that, if possible, should be avoided.
  • EO (Environmental Objections), where the review noted severe environmental impacts that should be avoided in order to properly protect the environment.
  • EU (Environmentally Unsatisfactory), where the review has determined that the environmental impacts are significant enough to warrant a statement regarding the project as something that should not continue as planned.
    
    For the numerical portion of the rating, there are three different categories:
  • The 1 rating indicates that the EIS is adequate and that no additional data collection is necessary, although some ideas on language changes and information may be provided.
  • The 2 rating indicates that the EIS has insufficient information. This rating indicates that there is not enough information contained within the text to make an adequate assessment of the impacts.
  • The 3 rating indicates that the project is inadequate. This goes as far as saying that the EIS does not meet the purpose of NEPA or the Section 309 review.
    
    The USEPA rating was generally found within the text of the EPA comment letter.
• Permits — Permits were a main area of focus within the data collection process, although the need for particular approvals were not always specifically stated in the EIS. Finding out whether a permit was needed for a certain project was often dependent upon the impacts of the Preferred Alternative. Each permit has a different set of guidelines that helped determine if it was necessary for the project or not.
  
  Section 404: Section 404 permits were the most commonly required permits for the universe of EISs sampled. The Section 404 permit is necessary when any jurisdictional waters of the United States are to be impacted. Section 404 details were found in the Permits section. The Permits section was usually found either in the Executive Summary or in the Environmental Consequences chapter. If there was no Permits section, then Section 404 permit information would be stated in or deduced from the Wetlands section of the Environmental Consequences chapter.
  
  Individual/Nationwide Permit: Individual and Nationwide are the two basic types of Section 404 permits issued by the ACOE. Individual and Nationwide permits are issued dependent on the size of the wetland impacted, nature of the action and the quality of the wetlands impacted. Individual Permits indicate a more severe impact while Nationwide Permits indicate a less severe impact. This variable was not actually included in the database, due to the difficulty in obtaining the information to differentiate, in most cases.
  
  Section 9 Bridge: A Section 9 Bridge permit is issued under the authority of U.S. Coast Guard. A Section 9 Bridge permit is required when the Preferred Alternative impacts navigable waters. Information regarding this permit is usually found within the Permits section of the EIS, if such a section is included. If there is no Permits section, it can usually be deduced from the Preferred Alternative section.
  
  Section 4(f): A Section 4(f) evaluation is required when a Preferred Alternative impacts any public park, recreation area, or wildlife / waterfowl refuge, or any significant historic site. Most EISs identify the need for a Section 4(f) on the Final EIS Signature Page. In addition, if a Section 4(f) evaluation was performed, it was usually included as a separate section within the Final EIS document.
  
  Section 106: Section 106 review is required if any National Register or National Register-eligible historical or archeological site is impacted. It is often not clearly stated in the text if Section 106 was required or not. When such a statement was not specifically included in the EIS, Section 106 review was assumed if a number greater than zero was entered into field 6.6 of the project data sheet (i.e., Number of Eligible or Potentially Eligible Historic/Archeological Sites Studied.
  
  Section 7: A Section 7 consultation is required if any federally threatened or endangered species are impacted within the project area. Similar to Section 106, it was unusual to find clear documentation of a Section 7 requirement included in the Final EIS. To determine if a Section 7 was needed, research was done within the Environmental Consequences chapter under the Threatened and Endangered Species section. If within the text there is specific language of any federally threatened or endangered species being impacted, then a Section 7 consultation was assumed.
  
  NPDES: This permit is necessary to help alleviate storm water runoff from impervious surfaces as they discharge into water resources. Information on whether a NPDES permit was required could be found within the Permits section, if such a section existed, or sometimes in the Executive Summary or the Water Quality section of the Environmental Consequences chapter.
  
  Other: This field is for any other state or federal permit that did not fall within the above fields. These permits were often listed in the Permits section, if such a section existed, or within the Water Quality section of the Environmental Consequences chapter.

• Environmental Issues/Controversies — A six­step methodology was implemented in determining whether a particular topic would be classified as an issue/controversy.
  • The first step was to read the Final EIS Signature Page. On the Final EIS Signature Page there was usually a brief description of the project along with an explanation of what the project entails. There were also a few sentences that summarized major adverse impacts. All major adverse impacts were considered issues/controversies except wetlands and displacements. These two impacts were looked at more closely before determining if the impacts for a particular project warrant them to be considered as an issue/controversy.
  • The second step was to read thoroughly through the Executive Summary. Contained in the Executive Summary was (usually) a brief look at all of the environmental impacts that would be involved with a certain project. Usually within the Executive Summary, a synopsis was given on a specific impact, which was usually a good indicator of what the issue/controversies would be.
  • The third step was to see if the EIS had a specific section for Controversies. If an EIS had this section it was usually found at the very end of the Executive Summary. Any impacts that were listed in this section were automatically considered issues/controversies.
  • The fourth step was to read the Environmental Consequences chapter for each individual impact. The Environmental Consequences chapter gave detailed information and mitigation strategies for all of the project impacts.
  • The fifth step was to read through the Public Involvement chapter. This chapter helped give an emotional pulse to the project. Individual letters received on the project from commenting / regulatory agencies as well as general public were reviewed for identification of specific issues or areas of controversy that were mentioned.
  • The final step was to take all of the above steps under consideration and decide which impacts warrant a label of issue/controversy. These determinations were generally based on the geographical size or magnitude of impacts, and the perceived significance of these impacts on the people and/or environment within the proposed project area. In some cases, the presence of an issue/controversy was identified in the comments received from agencies, public interest groups or general public. Below is a list of guidelines that were followed in order to help establish which issues qualified.
  •  Land use: All 4(f) properties warranted either a land use or cultural resource controversy. Any use of public land or Native-American land warranted a controversy.
  • Farmland: The three main factors used in determining if farmland was an issue/controversy were: how much farmland was being acquired, how much of that farmland was prime farmland, and how was the current farmland zoned. Farmland was determined to be an issue/controversy only when the impact was considered to be of direct economic and/or cultural significance to the people and/or community surrounding the impacted farmland. An exact acreage value could not be consistently applied as a threshold point for all projects since impact varied greatly from one EIS to another.
  • Economic/Fiscal Impacts: Economic and fiscal impacts become a controversy if the businesses and people employed by the businesses, within the project area, were economically hurt by a proposed project.
  •   Noise: Noise was always a factor when a roadway system increased its capacity. Deciding on whether it was an issue/controversy or not depended on how many homes were impacted and what type of setting the roadway was in. In urban settings, noise was of greater importance to the public than it was in most rural settings. Three major factors were looked at when making a decision on this topic: new decibel levels, amount of homes impacted, and amount of mitigation required. When considering noise, it was important to note that although mitigation was sometimes not feasible or reasonable it did not cancel out the fact that this impact could still be an issue/controversy.
  •   Air Quality: Air quality was often difficult to define as an issue/controversy because the air quality model used in the analysis generally showed improvement as a result of constructing the project. Such improvement was generally the result of reduced congestion and lowered emissions. Therefore, air quality was considered an issue/controversy only if an area was already in a non-attainment zone and getting worse, or the main opponents of the new roadway used the lessening of air quality as a major variable in their argument against a proposed Preferred Alternative.
  •   Visual Impacts: Defining whether visual impacts were an issue/controversy depended on the alignment of the Preferred Alternative and the type of setting it was in. Other large factors included: public opinion, and the viewshed affected.
  •  Wetlands: Impacts to wetlands were very common to most highway projects. Two factors where used when considering whether a wetland impact would be an issue/controversy. The first was the amount of acreage lost compared to total length of the project. The second was the quality of wetlands impacted. If either one of those factors were found to be lopsided than it was marked as an issue/controversy.
  •  Threatened and Endangered Species: Impacts to threatened and endangered species were considered to be a controversy when a direct impact to the existence of a species is threatened by a roadway project. Direct mitigation, which included the physical movement of a species or the rebuilding of a habitat, was a clear sign that this impact would be an issue/controversy.
  •   Cultural Resources: Cultural resources were determined to be an issue/controversy when a Preferred Alternative had an adverse effect on an historical or archaeological site that was on the National Register or considered to be eligible or potentially eligible for inclusion on the National Register. Most Native American impacts also fell within this category.
  •   Water Quality: Water quality was usually impacted in some form by roadway projects. Any time an impervious surface is added there is a chance for an increase in certain variables within waterways. However, the use of Best Management Practices and the regulations found within the NPDES permit mitigate much of the damage. Water Quality was generally considered an area of issue/controversy when the waterways impacted were used as part of the public water supply, or the waterway impacted was of the most pristine nature.
  •   Indirect/Secondary/Cumulative Impacts: Indirect/Secondary/Cumulative Impacts usually come under scrutiny when a roadway would directly help expand development. Determining whether this was an issue or not depended on whether or not new development was planned or zoned for, as well as taking into consideration the public's opinion on the matter.
  • Environmental Justice: Environmental justice impacts were based on the socioeconomic make-up of the area surrounding the project. If a project was proposed in a community that was characterized to consist of predominately low income or minority population, then environmental justice was considered to be an issue/controversy.
  •   Community Cohesion: Deciding on whether community cohesion was an issue/ controversy depended on the magnitude of the impact to the cohesion of a community. The magnitude and duration of new traffic patterns had to be taken into account. If existing travel patterns were anticipated to change as a result of the proposed project, the severity of the change and the loss of access to community centers, businesses and/or government buildings were taken into consideration. The severity of the impact was based on such factors as increased travel time, disruption of community travel patterns, segmentation of a community, magnitude of displacements, etc., and was considered in determining if community cohesion was an issue/controversy.
  •   Vegetation Impacts: To determine whether a vegetation impact occurred depended on where the project took place, the acreage impacted, and the ecological value of the area impacted. If the acreage impacted and/or the ecological value lost was high for that area, then vegetation was considered an issue/controversy.
  •   Business Displacements: To determine if this field was an impact, a couple of factors have to be examined. The first factor was how many businesses were displaced compared to the total number of businesses in the project area. The second factor dealt with the amount of adequate replacement property for these businesses. If there were a large number of displacements in comparison to the number of businesses in the project area, and/or a lack of space to relocate displaced businesses elsewhere within the project area, this became an issue/controversy.
  •   Residential Displacements: Residential displacements were a common occurrence for most projects. Therefore, to determine if this was an issue or not, certain guidelines were followed. The first guideline was to look at the number of displacements compared to the total number of residences within the project community. The second guideline was to see if replacement housing was prevalent within the project community. The third guideline was to see what kind of opposition neighborhood associations and other like groups put forth. If the number of homes displaced was high in comparison to the total number of homes in the project community, or if there was no replacement housing available, or if there was opposition from homeowners, residential displacements became an issue/controversy.
  •   Hazardous Materials: Hazardous materials were usually found within businesses, warehouses, etc. that were to be displaced or were abandoned within the Preferred Alternative´s right­of­way. If a substantial amount of hazardous materials or a small amount of a very dangerous material were found to be present, it was considered an issue/controversy.
  •   Floodplains: Floodplains were analyzed to see what types of encroachments would occur as a result of the project. Two factors were considered when identifying whether an issue/controversy existed: 1) the acreage of floodplains impacted, and 2) the history of the area for flooding. A sizeable encroachment and or a distinct history of flooding in the project area were the triggers for classifying Floodplains as an issue/controversy.
  •   Other: The "other" variable was reserved for any impacts that did not fall under any of the above-mentioned categories. Reasons for citing "other" as an issue/controversy were on a case­by­case basis.
• Number of Transit or System Alternatives Given Serious Consideration — In most cases, transit or system alternatives were considered and dismissed as being not feasible or not capable of achieving the identified purpose and need for a proposed project. Although many were considered, few were evaluated in detail for environmental impacts. Only those alternatives that were advanced for detailed analysis of environmental impacts were included in the total number given "serious consideration."
• Number of Design/Location Alternatives Given Serious Consideration — This was the total number of design/location alternatives that were given "serious consideration". This information was found in the Executive Summary or the Alternatives Chapter. The ones carried on for "serious consideration" were those looked at for "final consideration" in the EIS.
• Did the Preferred Alternative Change Significantly — This variable looked at whether the document made "significant changes" after a given stage in the EIS process (i.e., after the Draft EIS, a Supplemental Draft EIS and/or the Final EIS, in cases when a Supplemental Final EIS was prepared). Changes were identified by comparing the Preferred Alternative from each document stage and then cross referencing the Preferred Alternative maps provided at each stage. Determining if it was a significant change depended on the size and make-up of the change.
• Number of Households to be Relocated — This number was usually found within the Executive Summary and/or in the Environmental Consequences chapter as part of the Relocations section.
• Number of Businesses/Farms to be Relocated — This number, similar to residential impacts, was usually found within the Executive Summary and/or in the Environmental Consequences chapter as part of the Relocations section.
• Other Relocation Assistance to be Carried Out — This field includes any other relocation assistance that does not fall within the parameters of the Uniform Relocation Assistance and Real Properties Acquisition Act of 1970.
• Acres of Wetlands Impacted — This is the number of wetlands impacted by the Preferred Alternative. This number was usually given in the Executive Summary. However, if the acreage number was not in the Executive Summary, it could usually be found in the Environmental Consequences chapter as part of the Wetlands section.
• Acres of New Wetlands Created, Acres of Wetlands Enhanced/Restored — Finding an exact acreage amount for these two variables was often times very difficult. The acreage amounts, if found, would be within the Executive Summary, the Mitigation section or, more specifically, the Wetland Mitigation section. Most of the time there would be no acreage given, because the final mitigation plan would be developed during the design phase of the project.
• Other Wetland Mitigation Measures — Other wetland mitigation measures are any mitigation measures besides enhancement/restoration. Minimization techniques such as culverts and bridging were not considered in this variable.
• Miles of Noise Barriers — This number was seldom given within the Executive Summary. To find this number, the Environmental Consequences chapter within the Noise section, and the mitigation portion of this section were referenced. Often, instead of giving the actual miles of noise barriers proposed for the project, the information was relayed via maps. If the maps had a scale, a number would be obtained through measurement. If no scale was given NA was entered onto the data collection sheet.
• Number of Eligible or Potentially Eligible Historic/Archaeological Sites Studied — To render the number of potentially eligible historic/archaeological sites requiring study, an understanding of the Cultural Resources section in the Environmental Consequences chapter was obtained. Within this section there is text outlining impacts to historic and archaeological sites. However, it was imperative to make sure that only the sites that were considered eligible or potentially eligible for inclusion in the National Register were included in the count.
• Community Impact Mitigation Measures — Community Impact Mitigation Measures were often mentioned in the Executive Summary of an EIS. They were also discussed in greater detail in the Community Cohesion portion of the Environmental Consequences chapter. If no specific Community Impact Mitigation Measures were identified in these discussions, then it was assumed that no such measures were applied.
• Visual Impacts Mitigation — Similar to Community Impact Mitigation Measures, Visual Impact Mitigation measures were also found in the same two locations within the EIS.
• Other Major Mitigation Measures — Other Major Mitigation Measures used the same kind of methodology as Community Impact Mitigation Measures and Visual Impacts Mitigation Measures. In this regard, only those mitigation measures that were substantial or important in reducing/alleviating/counteracting impacts resulting from the project were included. Within a project, if an impact was being mitigated, it usually meant it played a role in the development of the Preferred Alternative. Using this line of reason, most mitigation measures were included in the Other Major Mitigation Measures field.
• Number of Public Meetings/Workshops, Public Hearings, Agency Meetings, Public Official Meetings — Ascertaining the total number of meetings varied from EIS to EIS. Each EIS approached revealing this information in a different fashion. Some EISs set up a chronological table with dates and documentation of every meeting. Other EISs used a summary format, and instead of numerical values, used catch-all words such as "many", "several" or "numerous." Other EISs would write of TAC meetings and CAC meetings that met monthly throughout the process. However, there was no way of telling when the first meeting date was, and the same could be said of the last meeting date. A numeric value of all meetings was used in this field, and included the sum of all of the individual types of meetings presented for the next several variables.
• Public Meetings/Workshops — This field included any meetings where project officials and the general public met for an exchange of ideas, or any meetings where information was updated by project officials to the general public. These sessions also allowed a chance for suggestions by the general public. A numeric value of all public meetings/workshops was used in this field.
• Public Hearings — Official hearings dealt with comments and issues raised within the Draft EIS. The hearings were governed over by project leaders and representatives of both FHWA and state DOTs. An official court reporter recorded all testimony given. Some EISs place the official transcripts in the public involvement section of the Final EIS. A numeric value of all public hearings was used in this field.
• Agency Meetings — These included any meetings that involved: TAC, PAC, Federal Agencies, State Agencies, Local Agencies, Neighborhood Coalitions, Activist Groups, and Local Companies. A numeric value of all agency meetings was used in this field.
• Public Official Meetings — These included any meetings that involved: Federally Elected Officials, State Elected Officials, City Councils, Community Councils, and Town Managers. A numeric value of all public official meetings was used in this field.
  

   2.4.2 Development and Verification of Database

The information from the project data forms was entered into a Microsoft Access database format that was suitable for statistical handling and analysis. A total of 87 separate data fields of information, reflecting the types of information included on the project data forms, were created from the outset. A list of the variable names assigned to each of the data fields used and their respective definitions is included in table 1. The actual final database containing 244 records (250 minus the six records that were eliminated as discussed in Section 2.3.2) and 87 fields per record is included in its entirety in Appendix D.

Whenever a project was missing data for a given variable, either because that variable didn´t apply to that project or because the answer could not be determined, the entry was generally coded with a "999" so that it would be excluded from any statistical analysis. The code "999" was used instead of "NA" (Not Available) since the variables were primarily numerical in nature and required a numerical value to indicate missing information as well. Once all of the data were entered into the final database, a visual quality control check was performed on the data to identify any anomalies such as milestone dates that didn´t make sense in relation to other dates. When necessary, the original data sources were consulted to verify that the data entered were correct. If a transcription error was identified, the entered data was corrected. If the data anomalies could not be explained or corrected, the data field entry was changed to "999" in order to delete the data from further analysis.

Back to Top

Source: The Louis Berger Group, Inc. 2002

Upon completion of the visual quality control check, a computerized check to further check for anomalies was conducted as described in Section 2.5.1 below. In cases where the missing data were in a variable considered to be critical for standardizing the database (e.g., Notice of Intent date), those records were eliminated from further consideration. Details of the project records eliminated from the originally-created 250 records in this manner are provided in Section 2.3.2.

Upon completion of the visual and computerized quality control checks, the database was ready for undertaking statistical analysis.

2.5 Statistical Analysis Methodology

2.5.1 Identification of Descriptive Statistics

Similar to the Phase I Study, descriptive parameters of the database variables were developed, first as a means to identify the general characteristics of each variable and then as a method of identifying any further anomalies in the data that were not caught during the visual quality control check. In this regard, the final database was entered into the SPSS Base 10 statistical software package, which allows a variety of descriptive statistical analyses to be performed.

The database was initially examined to determine if any values fell outside of a range of values that was considered to be appropriate for a given variable. Any values that suggested a potential error in the data or a suspicious outlier were identified and flagged for further scrutiny. Values for variables measured at the interval level were then examined in relation to the mean, standard deviation, skewness, range, and maximum and minimum values. Histograms, stem­and­leaf and box­and­whisker plots were created to provide a visual view of the data (examples of these types of graphic representations are provided in the Descriptive Statistics appendices). Probability plots, used to evaluate the degree to which interval variables vary from the normal distribution, were created for most continuous variables. In the case of nominal or ordinal variables, a frequency analysis usually sufficed to indicate any anomalies.

Next, a number of computed variables were created from the original 87 variables included in the final database presented in table 1. One of these variables, denoted as NEPATIME, was created to reflect the elapsed time (calculated in seconds) between the date of the NOI and the date on the signature page of the Final EIS. The resulting quantity was divided by 31,536,000 (24 hours x 60 minutes x 60 seconds x 365 days). The resulting number was a close approximation to the years and fraction thereof between the EIS start and end dates. This variable was designated NEPATIME and became the "raw" dependent variable for most of the analysis. It was the only variable created in this fashion.

In the above case, where one variable was subtracted from another, the results were checked to ascertain if an inappropriate value resulted. For example, if the year in which a process began was subtracted from the year in which it ended and a negative result occurred, it indicates that the end date was earlier than the begin date. An effort was then made to determine if there was a data entry error or if the source data was incorrect. In every case, the problem was corrected so that only those projects having positive NEPATIME values was ultimately included in the database. In addition, NEPATIME for one case was substantially longer than the others and represented a clear outlier. Consistent with the procedure used in Phase 1, cases above 15 years of EIS duration were eliminated from further analysis. As stated in Section 2.4.2 above, the final database contained 244 project records.

In addition to the new variable that was created to reflect the length of the NEPA process, some computed variables were created by adding one variable to another. These new variables included:

• TOTPERM = The sum of all permit and special studies variables for a given project, to a maximum of 7 (i.e. those variables listed between and including PERM404 and PERMOTHER in table 1).
• TOTISS = The sum of all issues variables for a given project, to a maximum of 19 (i.e., those variables listed between and including ISSULAND and ISSUHAZM in table 1).
• ALLREL = The sum of all relocations for a given project (i.e., total number indicated by the variables RELHH, RELBUS and RELOTHER, in combination, in table 1).
  

Some variables were transformed into different numeric equivalents such as square roots or logarithms. The purpose of such transformation was to convert interval variables that were not normally distributed into equivalents that approximate the normal distribution. The following new variables were created in this manner from either the initial variables listed in table 1 or other created variables:

• SQRTNTIM = the square root of NEPATIME (Length of NEPA Process in Elapsed Time)
• LGLENGTH = the log 10 of LENGTH (Length of the Project)
• CRTLNDAQ = the cube root of LAND_AQ (Acres of Land Acquired)
• LNALLREL = the natural log of ALLREL (All Relocations)
• LNWETACI = the natural log of ACWETLI (Acres of Wetlands Acquired)
  

The first of these became the transformed version of NEPATIME that would be used in procedures that required a variable that approximated the normal distribution. One problem with such transformations is that some values cannot be transformed and become missing values; e.g. the square root of 0. In a data set with a very large number of missing values to begin with, loss of further data is not desirable and where such variables were used, other tests were performed using the original data so as to minimize data loss.

A number of variables were reduced to categorical variables either because their structure suggested categories or because no attempted transformation succeeded in approximating a normal distribution. The categories were often based on quartiles, sometimes on location of the median and sometimes on more arbitrary grounds designed to avoid disproportionately large or small numbers of cases in the categories. Of course, some original variables, such as FHWAREG, MSA, GROWTH, and all of the permit and issues variables, as well as others were categorical. Created variables and their categorized responses in parentheses are presented below:

• CATCOOP = Number of Cooperating Agencies (COOP_AG in table 1), split into four categories: (0) 0; (1) 1; (2) 2 and 3; (3) 4 or more.
• CATCOM = Number of Commenting Agencies (COMM_AG in table 1), split into four categories: (1) 1—9; (2) 10—13; (3) 14—19; (4) 20 or more
• CATALT2 = Number of Design / Location Alternatives Given Serious Consideration (PROJALT2 in table 1), split into two categories: (1) 0—3; (2) 4 or more
• CATWETLR = Acres of Wetlands Enhanced / Restored (ACWETLR in table 1), split into two categories: (0) 0; (1) more than 0
• CATWETCR = Acres of Wetlands Created (ACWETCRE in table 1), split into two categories: (0) 0; (1) more than 0
• CATNOIS = Miles of Noise Barriers (NOISEB in table 1), split into two categories: (0) 0; (1) more than 0
• CATCRSIT = Number of Eligible or Potentially Eligible Historic / Archaeological Sites Studied (CRSITES in table 1), split into four categories: (0) 0; (1) 1—2; (2) 3—5; (3) 6—12; (4) 13 or more
• CATPUBM = Number of Public Meetings / Workshops (PUBLIC_M in table 1), split into four categories: (1) 0—2; (2) 3—4; (3) 5—6; (4) 7 or more
• CATAGME = Number of Agency Meetings (AGENCY_M in table 1), split into four categories: (1) 0—4; (2) 5—11; (3) 12—31; (4) 32 or more
• CATPUBH = Number of Public Hearings (PUBLIC_H in table 1), split into two categories: (1) 0—1; (2) more than 1
• CATPUBO = Number of Public Official Meetings (PUBLIC_O in table 1) split into four categories: (0) 0; (1) 1—2; (3) 3—7; (4) 8 or more
• NLCAT= Number of New Lanes Added (NEWLANES in table 1), split into four categories: (0) 0; (1) 1—3; (2) 4; (3) 5 or more
• CATNPATI = Length of the NEPA Process in Elapsed Time (NEPATIME as a created variable), split into four categories: (1) 0 — 3.1; (2) 3.1 — 4.7; (3) 4.7 — 6.6; (4) 6.6 or more
• GEOLOC 3 = Geographic Locations, derived from Former FHWA Regions (FHWAREG in table 1) and split into three categories: (East) Regions 1—4; (Central) Regions 5—8; (West) Regions 9 and 10
• GEOLOC = Geographic Locations, derived from Former FHWA Regions (FHWAREG in table 1) and split into two regions: (Coastal) Regions 1—4, 9 and 10; (Central) Regions 5—8.
  

Certain other variables were recoded to replace text values of the variables with numerical values. These are not obvious, since the numerical values were given text surrogates and appear as text even though they are actually numerical. These variables, which were derived from some of the initial variables presented in table 1, include the following:

• LUSE# = the numerical form of LAND_USE
• MSANUM = the numerical form of MSA
• GROW_NUM = the numerical form of GROWTH
• PROJTYP# = the numerical form of PROJTYP1
• RATE# = the numerical form of EPA_RATE

It should be noted that the addition of the created variables to the database increased the total number of variables to 116, although not all of those variables were used in the statistics analyses.

Interval variables, both computed and the original variables, were then run through the SPSS program´s Exploratory Data module, which repeated some of the descriptives, added others (e.g., medians, m-estimator measures, etc.) and consolidated much of the information into a single set of tables. An additional advantage of the Exploratory Data module is that it permits the compilation of data by categories, so that exploratory data could be generated by former FHWA Region and other categorical variables and then compare the results.

In developing the descriptive statistics, a number of assumptions were made. These included the following:

• All data were drawn from a normal population
• All cases are independent of each other
• Significance level was set at 0.05 and, in cases where the statistical package calculated the precise significance, it was provided.
• Assuming the data represent the entire population of available EISs, which should be the case, randomness is not an issue.

2.5.2 Selection of Data Analysis Methods

Descriptive and analytical methods were limited to procedures available in the SPSS Base 10 statistical package. The specific methods used from that package are described below.

  Identification of Data Distribution

Many statistical tests require that the data being analyzed be normally distributed. Although the Central Limits Theorem technically would not apply when an entire universe of cases is involved rather than a sample, an effort was made to either transform non-normally distributed data to a form approximating a normal distribution, or to test the data using procedures that do not require a normal distribution. In addition, certain procedures, while formally assuming normally distributed data, are in fact sufficiently robust to allow a departure from this assumption. Examples include the ANOVA procedure and the T-test procedure.

Interval data were examined for normalcy using both graphic and non-graphic methods. First, the variable was subjected to a Kolmogorov-Smirnov test against a theoretical normal distribution. If the significance level was greater than 0.05, the variable was assumed to not differ significantly from the normal distribution. If the significance level was 0.05, the variable was treated as being not normally distributed.

  Data Transformations

Data identified as being non­normally distributed were then visually examined using one or more histograms and probability plots to identify the most likely transformation method. If a given variable could be appropriately transformed using the techniques available in the statistical package, this was done. In some cases, the square root of the original variable resulted in a normal distribution. In other cases, the logn, log10, reciprocal or exponential transformations were tested. Transformation resulted in a new variable, but the original variable was retained. The key variables created through such transformation are identified above in Section 3.5.1.

  Parametric Versus Non­Parametric Analyses

Some variables in the final database could not be easily transformed. Others were clearly not interval data and were not normally distributed. For these variables, non­parametric test procedures were utilized. Unlike parametric procedures, non-parametric procedures do not require a normal distribution. Non-parametric tests are generally considered to have less power than parametric tests (i.e., there is a somewhat higher risk of a Type II error occurring, which involves the failure to reject a hypothesis when it is actually false). Since the norm in significance testing is to test a null hypothesis (e.g., there is no difference in the length of the NEPA process between Regions 1 and 5), the lower power of the non-parametric tests could result in failing to reject that hypothesis. Therefore, wherever possible, parametric tests were used. In each case, the transformed variable was tested parametrically and the original variable was tested non-parametrically to compare the results.

2.5.3 Statistical Relationship Test Between Variables

Statistical relationship tests were conducted in order to answer a number of questions, all of which relate to the length of the NEPA process. Specifically, these tests were designed to identify which variables, if any, vary with the length of the NEPA process, and which factors, if any, appear to differ with respect to the length of the NEPA process. A list of the relationships that were tested in this regard is shown in table 2.

The general approach to identifying any statistical relationships began by first looking at the descriptive statistics as a whole, and then by looking at the descriptive statistics broken down by categories suggested by the data variables themselves. For example, a total of 244 cases of NEPATIME information were available for analysis. These cases could then be grouped into categories (e.g., by FHWA region) and the descriptive statistics could be obtained by category (FHWA region) as well. The same process could then be followed for dividing the cases into, for instance, those associated with a Section 4(f) study and those not associated with a Section 4(f) study. Similar analyses could be performed for a variety of other types of categories as well.

Back to Top

Source: The Louis Berger Group, Inc., 2002

The next step was to attempt to find significant differences in the length of the NEPA process according to the categorical variables. The data for Phase II of the study differ from those of Phase I in one important way. In Phase I, the data were a sample drawn from a larger population.

It was appropriate, therefore to test the data to ascertain if two independent samples (e.g., cases in which a Section 4(f) procedure was involved and those for which it was not involved), were significantly different, that is, drawn from the same or different populations. It was also important to learn if the characteristics of the sample were likely to be the characteristics of the population.

In the present instance, it can be argued that the data set is the population, at least for the time period under study (i.e., 1995 — 2001). Apparent differences in mean values are what they are and it is not really necessary to test for significant differences. The counter argument is that since FHWA-related EISs existed prior to the time period under study and additional ones are being created and will continue to be created, the present data set is a time-defined sample of a larger population (i.e., the entire universe of FHWA-related EISs since implementation of NEPA through to the present).

In cases where the data are categorical and ordinal, it is useful to run an ANOVA using the SPSS Means procedure because, in addition to indicating whether or not subgroups exist within the population, the difference of means test (which involves an ANOVA) can show if the test variable varies with the categorical variable. It can also show the extent to which the ordinal categorical variable explains the variance of the test variable. That is to say, is there a correlation? What is its direction and what is its strength? It can also show if the correlation is linear or non-linear. These tests are only available where there are more than two categories and they are ordinally ranked. They produce both Pearson's R and Eta. The former requires a linear correlation, and its square shows the proportion of variance in the one variable that is explained by the other. Eta is analogous, but does not require a linear relationship. Its square indicates the proportion of variance in the dependent interval variable explained by the categorical variable. Eta may also be calculated for a nominal variable versus an interval variable in the SPSS Crosstabs procedure.

The means procedure was used for a number of categorical variables. The results are found in Appendix E.

For nominal categorical variables involving more than two categories, the SPSS one-way ANOVA procedure was used. The analysis of NEPATIME and SQRTNTIM by former FHWA region is an example of the group of analysis. The results of these ANOVAs and selected Post Hoc tests may be found in Appendix F.

T-tests were conducted for all dichotomous variables, not so much to ascertain if they are from different populations, but to assess the probability of obtaining differences in the dependent variable mean as large as those that were obtained. Exact significances are provided in these cases. The results of the various T-Tests may be found in Appendix G.

Finally, any correlations between the length of the NEPA process and other interval variables were sought. First, the two variables to be tested were placed in a scattergram. If a meaningful relationship between two variables existed, the pattern would be seen in the scattergram. The results of the scattergram were further confirmed either by obtaining Pearson´s R for approximately normal data or ascertaining the correlation through use of nonparametric procedures such as Spearman´s Rho. Spearman´s Rho also permits calculating a correlation between ranked categorical variables and between ranked categorical variables and interval variables.

A series of cross tabulations was also conducted using NEPATIME as a dependent variable cross tabulated with a series of categorical variables and using a categorized version of NEPATIME, CATNPATI. The Chi-square or variation of this statistic was utilized in the purely categorical crosstabs. Where a nominal categorical variable was cross-tabulated with an interval variable (NEPATIME), the Eta coefficient was calculated. In some cases it was appropriate to cross tabulate a nominal variable with an ordinal variable. In such cases the Chi-square statistic was relied on merely to ascertain if the variables are independent of each other.

Back to Top

3.0 RESULTS OF RESEARCH

This section presents the major results of the statistical analysis of this research. The first set of findings relate to descriptive statistics on the length of the NEPA process, which is really the heart of this entire analysis. The second set of findings relates to descriptive statistics on factors considered other than length of the NEPA process. The final set of findings relate to the relationship between length of the NEPA process and other factors considered.

3.1 Descriptive Statistics on the Length of the NEPA Process

The descriptive statistics on the length of the NEPA process for the country as a whole during the period 1995—2001 are presented first within this section, followed by descriptive statistics on the length of the NEPA process by the individual former FHWA regions.

3.1.1 Overall Descriptive Statistics

As discussed in the previous section, the Length of the NEPA Process was denoted in this analysis as the variable NEPATIME, which was calculated on the basis of the following:

NEPATIME = FEIS_SIG (the signature date of the Final EIS) - NOI (Notice of Intent date)

A descriptive analysis of NEPATIME was run. The results, which are presented in table 3 below and in both tabular and graphic form in Appendix H, indicate that the mean average value of NEPATIME during the 1995 — 2001 Phase II study period is approximately 5.1 years, with a standard deviation of 2.6 years and values ranging from 0.6 to 14.3 years. This mean value of NEPATIME is substantially greater than the mean of 3.6 years for the entire study period considered in Phase I. This is not surprising given that the Phase I mean of 3.6 years included mostly EIS projects completed during the 1970s and 1980s, when there were fewer specific regulatory requirements covered under the overall NEPA umbrella, and no EIS projects completed more recently than the early 1990s.

Back to Top

table 3:
LENGTH OF NEPA PROCESS (In Years)

Source: The Louis Berger Group, Inc., 2002.

Similarly, the NEPATIME mean value of 5.1 years for the 1995 — 2001 Phase II study period is substantially greater than the mean values of 2.2 years and 4.4 years for EIS projects completed in the 1970s and 1980s, respectively, and slightly higher than the mean value of 5.0 years for EIS projects completed in the early 1990s. Once again, this is not a surprising finding.

It should be noted, however, that the NEPATIME variable was found to be not normally distributed, indicating that the mean value may not be a true indicator of central tendency for that variable, so other indicators of central tendency were obtained. This was also the case in the Phase I analysis where the mean value was determined not to be a true indicator of central tendency. Upon further analysis, the median value, which is the midpoint among the full set of NEPATIME values that divides the observations into two groups of equal numbers, was determined to be about 4.7 years for the 1995 — 2001 Phase II study period, as indicated in table 3.

The reason that the median value is somewhat less than the mean value is because the higher values of NEPATIME that were included in the analysis (i.e., those approaching the highest value of 14.3 years considered in the analysis) tend to skew the mean on the high side, even though such higher values are in the minority. Therefore, in this case, the median value is likely to be a more realistic indicator of the typical length required for complying with the NEPA process during the 1995-2001 Phase II study period. This indication is confirmed by a series of so­called m­estimators generated by the SPSS statistical program. These m­estimators attach various weightings to ranges of values of a variable to arrive at an estimate of the actual central value that is more valid than the mean. In this case, the range of m-estimators was between 4.8 and 4.9 years.

The NEPATIME median value of 4.7 years for the 1995 — 2001 Phase II study period compares to the Phase I median value of only 3.0 years. Once again, this difference is not surprising given the difference in study years and the greater procedural and technical requirements imposed on the NEPA process in more recent years.

A Kolmogorov­Smirnov test comparing the distribution of NEPATIME with a theoretical normal distribution was also run. This test indicated that NEPATIME was significantly different from the normal distribution. A new variable, which was not significantly different from the normal distribution, was created by taking the square root of NEPATIME (SQRTNTIM). For many of the subsequent statistical analyses, SQRTNTIM was utilized instead of NEPATIME due to the more normal distribution of that variable, although equivalent NEPATIME values are also reported when appropriate. If one takes the mean value of SQRTNTIM (almost 2.2) and expresses it in terms of NEPATIME by squaring it, one finds that the mean SQRTNTIM is equivalent to almost 4.8 years in terms of NEPATIME. This is consistent with the m­estimators calculated for NEPATIME. Based on the mean and standard deviation produced by SQRTNTIM, it appears that most cases under study have a NEPATIME that lies between approximately 2.5 and 7.7 years.

Although the NEPATIME equivalent of 4.8 years is probably a better indicator of the true mean value for the Phase II dataset, it is appropriate to use the actual NEPATIME mean value of 5.1 years for purposes of comparison with the Phase I mean values. This is due to the fact that the mean values presented in Phase I reflected actual NEPATIME rather than NEPATIME equivalent.

The SQRTNTIM mean and median values, as well as SQRTNTIM standard deviation and inter­quartile range values are presented in table 3. Similar values for its NEPATIME equivalent are also presented in the table.

3.1.2 Descriptive Statistics by Region

The next step involved exploratory analysis of NEPATIME / SQRTNTIM categorized according to the nine former FHWA regions. The results are summarized in table 4 below, and are presented in greater detail in Appendix I. Those regions exhibiting the highest mean NEPATIME values are clustered close together, indicating that they have recently taken close to the same amount of time to complete the NEPA process on average, but have taken longer than any of the other regions. Region 4 (Southeast) demonstrated the highest NEPATIME value at 5.6 years, followed closely by Regions 1 (Northeast), 3 (Mid-Atlantic) and 10 (Northwest) at approximately 5.5 years each, and then by Region 9 (Southwest) at approximately 5.3 years. The median NEPATIME values for these five clustered regions are also higher than the other regions, although they are not necessarily in the same order as the mean values. The median values for these five clustered regions range from a high of 6.0 years for Region 3 (Mid-Atlantic) to a low of 5.2 years for Region 9 (Southwest).

Back to Top

table 4:
DESCRIPTIVE STATISTICS OF NEPATIME BY REGION

REGION/VARIABLE	CASES	NEPATIME		SQRTNTIM
		MEAN	MEDIAN	MEAN	EQUIVALENT
1 — Northwest	23	5.5	5.3	2.2543	5.1
3 — Mid­Atlantic	43	5.5	6.0	2.2491	5.1
4 — Southeast	33	5.6	5.7	2.3177	5.4
5 — Midwest	39	5.2	4.5	2.2023	4.8
6 — South Central	20	3.8	3.6	1.8948	3.6
7 — Plain States	22	4.4	3.9	2.0031	4.0
8 — Rocky Mtn. States	12	3.8	3.3	1.8766	3.5
9 — Southwest	34	5.3	5.2	2.2498	5.1
10 — Northwest	18	5.5	5.4	2.2796	5.2

Source: The Louis Berger Group, Inc., 2002.

At the other end of the spectrum, Regions 6 (South Central) and 8 (Rocky Mountain states) exhibited the lowest mean NEPATIME values of approximately 3.8 years, indicating that it has recently taken almost two years less to complete the NEPA process in those regions than in the longest regions. Their median NEPATIME values were also the lowest of all the regions, with values of 3.6 and 3.3 years for Regions 6 and 8, respectively.

On the basis of the NEPATIME mean and median values for all of the regions, some interesting comparisons with the results of the Phase I research can be made. Most interesting of all, perhaps, is the general magnitude of the values for the regions in this study in comparison to those from the Phase I study. Of the nine former FHWA regions, six of them are now shown to exhibit higher mean and median values than the highest region previously. Whereas Region 1 (Northeast) had the highest mean and median values of 4.7 and 4.5 years, respectively, in the Phase I study, the highest mean and median values of any region in the current study are 5.6 and 6.0 years, respectively, with the highest mean value in Region 4 (Southeast) and the highest median value in Region 3 (Mid-Atlantic). This fact is probably an indication of the ever-increasing complexity and difficulty in completing the NEPA process over time. The Phase I study only covered NEPA projects that were completed from inception of NEPA in the early 1970s through the early 1990s; in contrast, the current study only covers NEPA projects that were completed in the late 1990s and early 2000s. The general trend is similar to that for NEPA projects in all former FHWA regions combined, as presented in Section 3.1.1 above.

The other interesting finding in comparison to the Phase I study is that the particular regions that took the longest to complete the NEPA process during the 1970s through the early 1990s are still essentially those that have taken the longest during the most recent years. Regions 1 (Northeast) and 4 (Southeast), the two regions exhibiting the highest NEPATIME values in Phase I, still have two of the higher NEPATIME values although not the highest. The clustering of regions at the high end is probably an indication that other regions that formerly exhibited lower NEPATIME values have caught up to Regions 1 (Northeast) and 4 (Southeast) in terms of the difficulty in completing the process. This is especially true for Region 10 (Northwest), which had the lowest NEPATIME values in Phase I and now is among the highest.

As discussed previously in Section 3.1.1, the mean NEPATIME values presented in table 4 are probably somewhat overstated due to the fact that they were found to be not normally distributed. Therefore, the median values could be considered better indicators of the central value of NEPATIME than the mean values. Even better indicators are the SQRTNTIM mean values and, using those values, developing NEPATIME equivalents of the mean values. These are better indicators because the SQRTNTIM variable was found to be more normally distributed than NEPATIME.

Using the SQRTNTIM / NEPATIME equivalent mean values as presented in table 4, the results are fairly consistent with those found using the basic NEPATIME values, although they are somewhat lower. Region 4 (Southeast) still demonstrated the highest NEPATIME (equivalent) value at 5.4 years, followed by Region 10 (Northwest) at 5.2 years and Regions 1 (Northeast), 3 (Mid-Atlantic) and 9 (Southwest) at approximately 5.1 years. Similarly, at the other end of the spectrum, Regions 8 (Rocky Mountains states) and 6 (South Central), still demonstrated the lowest NEPATIME (equivalent) values at 3.5 and 3.6 years, respectively.

As was the case for the entire dataset of all regions combined, the NEPATIME equivalent mean values are somewhat better indicators of the true mean values for the individual regions studied in Phase II. However, it is once again appropriate to use the actual NEPATIME mean values for the individual regions in Phase II for purposes of comparison with the mean values for the individual regions in Phase I. This is also due to the fact that the mean values presented in Phase I reflected actual NEPATIME rather than NEPATIME equivalent.

Further geographic analysis was also performed in an effort to obtain any additional trends that may exist. A cross tabulation of the variables FHWAREG (the former FHWA regions) with CATNPATI (a categorized version of NEPATIME) provided some interesting geographical differences in the frequency of cases in each region to upper and lower halves of NEPATIME range. For example, Regions 1 (Northeast), 3 (Mid-Atlantic) and 4 (Southeast) had approximately 52, 60 and 70 percent of their respective cases in the upper half of the NEPATIME distribution. Conversely, Regions 5 (Midwest), 6 (South Central), 7 (Plains states) and 8 (Rocky Mountain states) had 59, 75, 64 and 82 percent of their cases, respectively, in the lower half of the NEPATIME range. Then, similar to the distributions for the three eastern regions, Regions 9 (Southwest) and 10 (Northwest) had 59 and 56 percent of their cases, respectively, in the upper half of the NEPATIME range. The Cramer's V had a significance of 0.01, indicating that the regions and the quartiles of NEPATIME are not independent of each other, although the low value of V (0.24) indicates a weak relationship. The Eta value generated by examining NEPATIME by region finds that only about 5 percent of the variance in NEPATIME is accounted for by location within various FHWA regions. Although the ANOVA that tested the difference in the means of the dependent variables by each of the former FHWA regions did not find significant differences at the .05 significance level, the actual significance level was approximately 0.08. The individual regions were then collapsed into three categories in a variable called GEOLOC3. Regions 1 - 4 are identified as "East" Regions 5 - 8 are identified as "Central" and Regions 9 and 10 are identified as "West. An ANOVA shows that there is a difference among the three categories with respect to the mean SQRTNTIM, with a significance of 0.02. Post hoc tests show that Category "East" forms one homogeneous group and Region "Central" forms another. Category "West" could belong to either group at the 0.05 significance level, but is clearly coupled with "East" if the test significance level is raised to 0.10. If the regions are collapsed into "Coastal" (Regions 1, 3, 4, 9 and 10) and "Central" (5, 6, 7 and 8) categories, a T­test of the Coastal/Central variable GEOGLOC * SQRTNTIM shows a significance of 0.00. The mean SQRTNTIM for "Coastal" regions was approximately 2.3 (NEPATIME equivalent = 5.1 years) and for "Central" it was approximately 2.0 (NEPATIME equivalent = 4.2 years).

These results, which are presented in table 5, indicate that it may be more appropriate to consider geographical differences in the length of the NEPA process by geographical categories that are larger than the former FHWA regions. In the case of the three categories, "East," "Central" and "West," it appears that there may be a real difference in the amount of time required to comply with NEPA, with the greatest time required in the "East" and "West" areas. This finding is consistent with the analysis of "Coastal" and "Central" categories, where the Coastal areas were found to require the most time to comply.

Back to Top

Source: The Louis Berger Group, Inc., 2002.

3.2 Descriptive Statistics on Other Factors Considered

Similar to the NEPATIME variable, descriptive statistics were also analyzed for the other factors to be tested for statistical relationships with NEPATIME and/or its more normally distributed square­root value, SQRTNTIM / NEPATIME equivalent. Highlights of selected descriptive statistics, which include number of valid cases and frequencies for categorical variables, as well as number of valid cases and mean, median, standard deviation and inter-quartile range values for continuous variables, are presented in table 6. The particular variables presented in the table include those which were subsequently found to have significant statistical relationships with NEPATIME and/or SQRTNTIM / NEPATIME equivalent, as well as some others considered to be key variables which are not presented in detail elsewhere in this section.

Back to Top

Note: NA indicates "Not Applicable."
Source: The Louis Berger Group, Inc., 2002.

Frequencies for all categorical variables tested, including some not shown in table 6, are presented in Appendix J. These frequencies indicate the number of responses for each categorical response that were received within the sample. For instance, in the case of the LAND_USE variable, 115 out of the 244 valid responses were identified as being rural in character, while 98 were identified as being suburban and 31 were identified as urban. Similar frequencies are presented in Appendix J for additional variables as well. These frequencies were then used to identify the appropriate types of further analyses to be performed (i.e., whether to treat them as continuous or categorical variables in the subsequent testing).

3.3 Statistical Relationship of the Length of the NEPA Process with Other Factors

Having developed the descriptive statistics for the length of the NEPA process and other factors, the next step was to refine the search for correlations and relationships between NEPATIME (or its surrogate) and other variables in the data set. The surrogate variable for NEPATIME, SQRTNTIM, which represents the square root of NEPATIME, was used when the tests involved indicated an assumption of a normal distribution. Since the interval variables in the present dataset were not normally distributed and many were not easily transformed to a near­normal form, non­parametric tests were often used. When this was the case, relationships were generally tested both with NEPATIME and with its surrogate. The reason that SQRTNTIM was generally used in identifying linear correlations using Pearson´s "R" correlations was because it more closely approximates a normal distribution than NEPATIME.

After visual examination by means of a scattergram for most interval data variables, Pearson (parametric) or Spearman (nonparametric) correlation tests were run against NEPATIME and SQRTNTIM. Ranked (ordinal) categorical data were compared with NEPATIME and SQRTNTIM to see if the dependent variable values varied with the categories. The SPSS Means procedure is appropriate for use in making such comparisons because it offers an optional ANOVA which produces an estimate of the correlation between categorical variables and continuous dependent variables such as NEPATIME or SQRTNTIM.

Unlike in the earlier Phase I research where a variety of other factors were found to have a statistical correlation or relationship with the length of the NEPA process, few correlations or relationships were found to exist in this follow-up study within the 0.05 level of significance. In fact, only two significant correlations with interval variables were found in this study, and both were rather weak. One of those cases is relatively unexplainable as to why such a correlation even exists. In addition, significant relationships between three categorical variables and NEPATIME / SQRTNTIM were found to exist as well. Once again, the connection between those variables and NEPATIME / SQRTNTIM is not clear or easily explainable.

The reason for the current lack of correlations and/or relationships between key factors that were previously found to have influence on the length of the NEPA process (e.g., Section 404, Section 4(f), Noise issues, etc.) and the dependent variables is unknown. Perhaps it is indicative of the fact that no single factor can explain the length of the NEPA process in recent years, as projects have become even more complex and other factors have become more interrelated than in earlier years. Whatever the reason is, the fact remains that only a few of the many factors considered can be used to explain the length of the NEPA process in a statistical fashion.

Of all of the relationships tested, only two significant correlations were found between tested interval or continuous variables and the dependent variables. One was between COMAGSTA (the number of state agencies commenting on the EIS) and NEPATIME. Although the correlation between these two variables is statistically significant and the probability of its occurrence by chance is small, the strength of the correlation between the two variables is weak. The correlation between these two variables as determined by the Spearman method and the correlation coefficient, Rho, was 0.18 with a significance of 0.00 and n = 242. These results indicate that little of the variance in NEPATIME is explained by COMAGSTA. From a practical perspective, it is also unclear why a relationship exists between these two variables, although the number of state agencies commenting may be indicative of the level of complexity of the EIS project.

Utilizing the Pearson and Spearman methods of correlation, only one other interval variable was found to correlate to either of the dependent variables at the 0.05 confidence level. The second significant correlation identified was between RATE# (numerical form of the EPA ratings variable, EPA_RATE) and NEPATIME, utilizing the Spearman Rho and Kendall´s tau_b. EPA ratings were converted to a numerical form and stored in a variable called RATE#. The numbers ranged from 1 to 11, with 1 representing the least EPA level of concern (LO­1) and 11 representing the most EPA level of concern (EU­3). Labels were then assigned to the numbers so that they would appear as the original ratings. It should be noted that there are more than eleven possible EPA rating combinations, but only eleven existed within the research database. RATE# was then correlated to NEPATIME using both Spearman Rho and Kendall´s tau_b. Both tests demonstrated a significant but weak positive correlation. Kendall´s tau_b presented a correlation of 0.14 with a significance of 0.01. The Spearman Rho produced a correlation of 0.18 with a significance of 0.01. In both cases, n = 204.

The relationships between a variety of other variables and the dependent variables, NEPATIME and / or SQRTNTIM, were also tested, using T­tests for the dichotomous or categorical variables versus SQRTNTIM. As stated previously, the purpose of these T­tests was not to determine if the values were derived from different populations, but to estimate the probability of differences of the size obtained. Only three variables showed differences in the mean values of SQRTNTIM that were significant at the 0.05 level:

VARIABLE		SIG.
ISSUECON	(Indicator of whether or not economic/fiscal impacts was a controversial issue)	0.02
ISSUWQ	(Indicator of whether or not water quality was a controversial issue)	0.01
CATPUBH	(Categorized version indicating number of public hearings)	0.03

Another variable, CHGDEIS (Indicator of whether or not the Preferred Alternative changed substantially after the DEIS), exhibited a significance level of 0.06, which was just slightly over the 0.05 level used in this study. Due to its proximity to the threshold level, it was determined that variable merits further consideration as a factor affecting NEPA process time.

Although a number of other significant correlations and/or relationships were also detected in the tests, none of them involved either of the dependent variables, NEPATIME or SQRTNTIM. Since these correlations did not involve a variable indicating the length of the NEPA process, they were not of specific interest to this study. Even so, a few examples of the other significant correlations and/or relationships are presented below, while the full array of correlations and relationships are presented in the matrices in Appendix K:

 LGLENGTH (Log10 of LENGTH) * CRTLNDAQ (Cube root of LAND_AQ)
 LNALLREL (Natural log of ALLREL) * LGLENGTH (Log10 of LENGTH)
 ALLREL (Casewise sum of RELHH, RELBUS and RELOTHER) * LAND_AQ (Acres of land acquired for the project)
 CRSITES (Number of eligible or potentially eligible historic / archaeological sites) * TOTPERM (Casewise sum of all permit and special studies variables)

The results of all correlation tests between the various factors considered and the dependent variables indicating length of the NEPA process are described below, regardless of whether or not they were found to be statistically significant correlations. These results are presented according to general variable types:

• Geographical Setting Variables — The nominal variables LUSE# (Numerical form of LAND_USE), MSANUM (Numerical form of MSA) and FHWAREG (Former FHWA regions) were tested against NEPATIME using the Crosstabs procedure. None showed an Eta value that demonstrated a significant relationship with NEPATIME.
  
  GROW_NUM (Numerical form of GROWTH), used as a ranked variable, was compared with NEPATIME and SQRTNTIM using the SPSS Difference of Means procedure. The ANOVA showed no significant difference among the GROWTH categories. Tests for linearity showed no linear relationship and no clear non-linear relationship. Eta, which does not require a linear relationship and which predicts the variation in the interval variable due to the categorical variable, suggests that GROWTH explains only two percent of NEPATIME variance.
• Project Type Variables — NLCAT (NEWLANES placed into four categories), used as a ranked variable, was compared with NEPATIME and SQRTNTIM using the SPSS Difference of Means procedure. The ANOVA showed no significant difference among the NLCAT categories. Tests for linearity showed no linear relationship and no clear non­linear relationship. Eta, which does not require a linear relationship and which predicts the variation in the interval variable due to the categorical variable, suggests that NLCAT explains only 1.3 percent of NEPATIME variance.

• Project Alternatives Variables — No significant relationships between the dependent variables and the number of project alternatives were identified. The dependent variables were not tested against PROJALT1 (Number of transit system alternatives given serious consideration) because of the large number of projects not considering such alternatives (212) and those that did (32). PROJALT2 (Number of design / location alternatives given serious consideration) was only somewhat more evenly distributed, but considered to be suitable for a correlation test using Spearman´s Rho. The results, however, showed no significant relationship with the dependent variables.
• Permit and Special Approval Variables — A number of permit and special approval requirements demonstrated sizeable differences in the length of the NEPA process, reflecting whether or not such permit / approval was required for a given project. Although none of these variables were found to be statistically related to NEPATIME / SQRTNTIM within the 0.05 confidence level, the fact that such sizeable differences in the length of the NEPA process were found between those projects where particular permits and/or special approval processes were involved and those projects where such permits and/or special approval processes were not involved, it is worth considering these relationships. Specifically, it is useful to examine both the mean and the median differences among cases requiring particular permits or approvals and those that did not.
  
  These variables are all dichotomous nominal variables indicating whether the case required the particular permit or approval ("True"), coded as a "1", or did not ("False") coded as a "0." These are presented in their entirety in Appendix K and are summarized in table 7.
  
    In considering the mean values for NEPATIME, it was found that a Coast Guard Bridge permit may extend the length of the NEPA process by more than seven months. Whereas the mean NEPATIME value is 5.6 years for projects involving a Coast Guard Bridge permit, those projects where no such permit was required demonstrates a mean NEPATIME value of only 5.0 years. In the case of the NEPATIME equivalents derived from SQRTNTIM, the difference between those projects involving a Coast Guard Bridge permit and those not involving such a permit is only about six months, with the actual mean values being somewhat less than the regular NEPATIME mean values (i.e., 5.2 years and 4.7 years for those with and without the permit requirement, respectively). In either case, however, a sizeable difference is indicated. It should be noted that the number of projects requiring a Coast Guard permit during the 1995 — 2001 period has been relatively small (197 "False" responses versus 47 "True").
  
  Section 4(f) evaluations also exhibited a sizeable, albeit not statistically significant difference between those projects involving such evaluations and those not involving such evaluations. In the case of mean values for NEPATIME, approximately six months difference exists between the two groups of projects, with those requiring the Section 4(f) evaluation being the higher of the two (i.e., 5.4 years vs. 4.9 years). In the case of the mean values for the NEPATIME equivalent derived from SQRTNTIM, less than five months difference between the two groups is indicated. The mean values for the NEPATIME equivalent are also somewhat lower in comparison to the actual NEPATIME mean values as well (i.e., 5.0 years and 4.6 years for those with and without the Section 4(f) evaluation required, respectively). The distribution of "True" and "False" responses for this variable is more even than in the case of the Coast Guard Bridge permits (i.e., 95 "True" responses and 149 "False" responses).

Back to Top

• Source: The Louis Berger Group, Inc., 2002.  
  
  Surprisingly, the other permit / special approval that showed a sizeable difference in the length of the NEPA process between those projects with and without such approval is NPDES. As with the Coast Guard Bridge permit and the Section 4(f) evaluation variables, the relationship between NPDES and any of the variables indicating length of the NEPA process is not significant to the 0.05 level. Those projects requiring a NPDES approval were found to exhibit a mean NEPATIME value of 5.3 years while those projects not requiring a NPDES approval were found to exhibit a mean NEPATIME value of 5.0 years, a difference of between 3 — 4 months. For the NEPATIME equivalent derived from SQRTNTIM, the 3 — 4 month difference still applies, although the actual mean values are less (i.e., 5.0 years for projects requiring NPDES and 4.7 years for projects not requiring NPDES). The distribution of "True" and "False" responses for this variable is fairly even, with 100 "True" responses and 144 "False" responses.
  
  It should be noted that for each of the above three permit / special approval variables, the median NEPATIME values were found to exhibit only a minor difference between those projects requiring the approval and those projects not requiring the approval. The median NEPATIME values for all three variables, including those with and without the various permit / special approvals, are somewhat lower than the corresponding mean values. In some cases, these median values reflect NEPA process times that are as much as 7 — 10 months shorter than when the mean values are considered. In all three cases, the median NEPATIME differences between those projects requiring a particular permit / special approval and those not requiring such permit / special approval are less than when the mean values are considered.
  
  In the case of Section 7 consultation, Section 106 approval and Section 404 permitting, it was found that virtually no difference exists in the mean values of NEPATIME or the NEPATIME equivalent derived from SQRTNTIM. The Section 404 finding, in particular, is surprising and impossible to explain, given that a sizeable and statistically significant difference in the length of the NEPA process was found between those projects with and without the permit requirement during the Phase I research. When analyzing the median values of NEPATIME associated with all three of these permit / special approval variables, the relationship becomes even more obscured given that two of them (i.e., Section 7 and Section 404) exhibit higher median values for projects not involving those particular approvals. Once again, since no significant relationships to the 0.05 confidence level were found for any of these three variables, the findings are likely to be meaningless.
• Issues Variables — The issues variables related to whether or not specific issues were identified in the EIS. These variables were of the same nature and were dealt with in the same manner as the permit and special approval variables, above. Some of the issues variables showed remarkable differences in the length of the NEPA process, depending upon whether or not the issue was present in the EISs. These large differences are noted in table 8 below for economic issues (ISSUECON) and water quality issues (ISSUWQ). As stated previously, these are two of the few variables to demonstrate a statistically significant relationship to the 0.05 confidence level with a variable indicating the length of the NEPA process.

Back to Top

table 8:
ISSUES VARIABLES AND THE LENGTH OF THE NEPA PROCESS

	Cases	Mean NEPATIME (years)	Median NEPATIME (years)	Mean SQRTNTIM	Median SQRTNTIM	NEPATIME Equivalent (years)
ISSUECON
True	31	6.1	5.9	2.4159	2.4304	5.8
False	213	5.0	4.6	2.1505	2.1537	4.6
ISSUWQ
True	77	5.9	6.2	2.3371	2.4960	5.5
False	167	4.8	4.5	2.1137	2.1171	4.5

Source: The Louis Berger Group, Inc., 2002.

The difference between the mean NEPATIME for cases for which ISSUECON is "True" and those for which it is "False" is more than a year. Using the median NEPATIME, the difference is more than 15 months. Using the mean NEPATIME equivalent derived from SQRTNTIM, the difference is about 17 months. Similarly, the difference between the mean NEPATIME for cases for which ISSUWQ is "True" and those for which it is "False" is also more than a year. Using the median NEPATIME, the difference is about 21 months. Using the mean NEPATIME equivalent derived from SQRTNTIM, the difference is just over a year. The disproportionately small number of cases coded as "True" may have some bearing on the differences in the values reported.

Unlike the permit and special studies variables, the pattern is not consistent among the issues variables. That is, many of these variables show longer NEPA process times where a particular discipline was not identified as an issue in the EIS. For example, for the flood plain variable (ISSUFLOODP), the NEPA process time was found to take between 6 and 9 months longer when there is no floodplain issue involved. Several other issues variables follow this same example. This outcome, in reality, is not likely and can be explained, at least partially, by the fact that these issues variables were not found to have statistical relationships with SQRTNTIM that are significant to the 0.05 level.

• EIS Change Variables — There were three variables in this category; namely, CHGDEIS (Indicator of whether or not the Preferred Alternative changed substantially after the DEIS), CHGSDEIS (Indicator of whether or not the Preferred Alternative changed substantially after the SDEIS) and CHGFEIS (Indicator of whether or not the Preferred Alternative changed substantially after the FEIS). Of these, only CHGDEIS was distributed in such a way as to make any analysis meaningful (189 "No" to 55 "Yes" responses). The other two variables were distributed 234 to 10 and 240 to 4, respectively. CHGDEIS showed no significant relationship to either dependent variable at the 0.05 level, but as explained above, the actual significance level of 0.057 indicates a low enough probability of finding so large a difference in the means, that further investigation might be warranted. Cases requiring a substantial change of the Preferred Alternative following the DEIS had a mean SQRTNTIM of 2.3166 (NEPATIME equivalent = 5.4) versus 2.1457 (NEPATIME equivalent = 4.6).
  

• Relocation Variables — This group of variables included RELHH (Number of households relocated), RELBUS (Number of businesses relocated) and ALLREL (Number of all households and businesses relocated). Also considered was RELOTHER (Other relocations), which was simply a "Yes" / "No" response variable. No significant correlation or relationship between any of these variables and any length of NEPA process variables at the 0.05 level was found using a T­test.
  

• Cultural Resources Sites Variables — No correlation was found between CRSITES (Number of eligible or potentially eligible historic / archaeological sites studied) and NEPATIME utilizing the Spearman method. Likewise, using the Difference of Means procedure, the ANOVA showed no significant differences among the categories of CATCRSIT (the categorized version of CRSITES) in the dependent mean values. Eta was only 0.116 for SQRTNTIM and 0.123 for NEPATIME.

• Mitigation Measures Variables — T­tests of the mean SQRTNTIM values for dichotomous mitigation variables were run and no differences in mean values were significant at or lower than a confidence level of 0.05. A T­test of CATNOIS (the categorized version of NOISEB, or miles of noise barriers) against mean values of SQRTNTIM showed no significant differences in the mean values for the length of the NEPA process.
  
  The variable, COMTYMIT (Community impact mitigation measures), was the closest of all of the mitigation variables to being significant, with a significance of 0.07. The mean SQRTNTIM values for cases with COMTYMIT mitigations was 2.2890 (NEPATIME equivalent = 5.2 years) versus 2.1412 (NEPATIME equivalent = 4.6 years) for those without.
  

• Meetings and Commenting Agencies Variables — Meetings of various types accounted for four variables; namely, Public_M (Number of public meetings), Public_H (Number of public hearings), Agency_M (Number of agency meetings) and PUBLIC_O (Number of public official meetings). None of these variables correlated strongly at the 0.05 significance level with the dependent variables. Categorized versions of three of these (CATPUBM, CATAGME and CATPUBO) were also tested, using the Means procedure. None showed a significant difference among its group categories for either NEPATIME or SQRTNTIM. Categorized forms of all four variables were tested against CATNPATI (categorized version of NEPATIME) for the Chi­square statistic, which was found to be not significant at the 0.05 significance level.
  
  For the dichotomous variable, CATPUBH (categorized version of Public_H), a T­test was run against SQRTNTIM and the probability of a difference of means as large as the one obtained is only about 0.03, thereby making the relationship significant to the 0.05 confidence level. As stated previously, CATPUBH is one of only a few variables tested in this study that were shown to have a significant statistical relationship with the length of the NEPA process. Cases requiring 0­1 public hearings were found to have a SQRTNTIM mean of 2.1158 (NEPATIME equivalent = 4.5 years) in comparison to 2.2917 (NEPATIME equivalent = 5.2) for those requiring more than one hearing. Despite the significant statistical relationship, it is not entirely clear how that relationship really explains the effect that the number of public hearings has on the length of the NEPA process, except that the number of public hearings may be an indicator of complexity of the particular project and/or the number of communities affected by the project. The number of public hearings may also be affected by whether or not a supplemental EIS process was involved. In any of these situations, it is reasonable then to conclude that those projects exhibiting a greater number of public hearings could result in a longer NEPA process time.
  As stated earlier, one other variable that was found to have a statistically significant correlation with the length of the NEPA process to the 0.05 confidence level was COMAGSTA (the number of state agencies commenting on the EIS). Although tests for correlation were also made for variables indicating the number of federal and local commenting agencies, as well as total commenting agencies, the only one that was significant was the state agencies variable. Although statistically significant, the correlation between COMAGSTA and NEPATIME was found to be rather weak. As a result, the ability to explain the relationship is also rather weak. It is possible, once again, that the correlation may be related to the fact that the number of state agencies commenting on a particular project may be indicative of the overall complexity of the project.
  

• EPA Rating Variable — Only one EPA rating variable was considered, which was RATE#, or the numerical form of EPA_RATE. As discussed previously, a significant but weak positive correlation was found when correlated with NEPATIME. Although this correlation indicates that the least severe ratings could lead to a shorter length NEPA process and that the most severe ratings could lead to a longer length NEPA process, the relationship is so weak that it wouldn´t prove to be very useful or conclusive in making such a conclusion.
  

3.4 Use of Results as a Baseline for Evaluating Future Environmental Streamlining Initiatives

Given the limited number of statistical correlations and relationships that were found to be significant to the 0.05 confidence level, as well as the sometimes unexplainable reasoning as to why those particular correlations and relationships have been shown to exist, it continues to be premature on the basis of the present data and findings to attempt to construct a predictive model of the length of the NEPA process. Although the data lends itself well to be used as a baseline against which to evaluate the relative degree of success of future environmental streamlining initiatives, its usefulness as a predictive model for determining the length of the NEPA process based on certain project or environmental factors is limited at this point in time.

As with the Phase I findings, the results of this Phase II effort can easily serve as a baseline from which additional analysis can be initiated and to which additional data can be added. These results also continue to expand upon the extensive FHWA database used for developing a historical baseline regarding the length of time required for completing the NEPA process, at least from Notice of Intent date to signature date of the Final EIS or Supplemental Final EIS, as applicable. The strongest findings of this study are perhaps related to the establishment of general descriptive statistics, including those presented by geographic location. The findings in this regard are also generally compatible and explainable with the findings of the Phase I research, in that the two sets of data and conclusions work together to tell an entire story about how the length of the NEPA process has continued to change over time. Therefore, the results serve as an excellent baseline for evaluating future environmental streamlining initiatives.

Back to Top

4.0 SUMMARY AND CONCLUSIONS

4.1 Summary of Findings

As presented in Section 3.0, this research has resulted in a number of interesting, as well as number of less interesting findings related to the length of the NEPA EIS process. A summary listing of the major findings is presented below.

• There were 250 FHWA projects resulting in signed Final EIS or Supplemental Final EIS documents pursuant to NEPA that were found to comprise the study universe of EIS projects during the 1995 — 2001 period.
• Based on the 1995 — 2001 study period, the average length of time for preparing and completing an EIS pursuant to NEPA was either 5.1 years or 4.8 years, depending on whether the mean value used was based on actual NEPA process times or equivalent NEPA process times derived from the more normally distributed square root values.
• Based on the 1995 — 2001 study period, the median length of time for preparing and completing an EIS pursuant to NEPA was 4.7 years.
• The mean length of time to complete an EIS during the 1995 — 2001 Phase II study period was substantially greater than that for EIS projects completed during the period from the early 1970s to the early 1990s, which was the period studied during Phase I. The mean of 5.1 years for the Phase II study period, based on actual NEPA process times, compares to only 3.6 years for the Phase I study period.
• The mean length of time to complete an EIS during the 1995 — 2001 Phase II study period was greater than that for EIS projects completed during each separate decade studied during Phase I. The mean of 5.1 years for the Phase II study period, based on actual NEPA process times, compares to 2.2 years in the 1970s, 4.4 years in the 1980s and 5.0 years during the early 1990s.
• The median length of time to complete an EIS during the 1995 — 2001 Phase II study period was greater than that for EIS projects completed during the 1970 — early 1990s Phase I study period, with a median value of 4.7 years in Phase II in comparison to only 3.0 years in Phase I.
• Former FHWA Region 4 (Southeast) exhibited the highest mean value (5.6 years based on actual NEPA process times, or 5.4 years based on equivalent NEPA process times derived from the more normally distributed square root values) related to length of time to complete the EIS process during the 1995 — 2001 study period. Clustered together close behind were Regions 10 (Northwest), 1 (Northeast), 3 (Mid­Atlantic), and 9 (Southwest), in approximate descending order.
• Former FHWA Regions 8 (Rocky Mountain states) and 6 (South Central) exhibited the lowest mean values related to length of time to complete the EIS process during the 1995 — 2001 study period. The mean value for both regions, based on actual NEPA process times, was 3.8 years while the mean values based on equivalent NEPA process times derived from the more normally distributed square root values varied from 3.5 years for Region 8 to 3.6 years for Region 6.
• The same former FHWA regions exhibiting the highest mean values related to length of time to complete the EIS process during the 1995 — 2001 period also exhibited the highest median values, but not necessarily in the same order. Similarly, the two former FHWA regions exhibiting the lowest mean values also exhibited the lowest median values.
• Of the nine former FHWA regions, six of them were shown to exhibit higher mean and median values during the 1995 — 2001 study period than the highest region during the 1970 — early 1990s Phase I study period.
• Regions 1 (Northeast) and 4 (Southeast), the two regions exhibiting the highest mean values related to length of time to complete the NEPA process during the 1970 — early 1990s Phase I study period, are still among the cluster of regions that have taken the longest during the 1995 — 2001 period.
• Geographical differences in the length of time to complete the NEPA process can and probably should be considered on the basis of "East," "Central" and "West" regions, with the longest times taken in the "East" region, followed closely behind by the "West" region during the 1995 — 2001 study period. The mean value for the highest region was 5.5 years, based on actual NEPA process times, or 5.2 years, based on equivalent NEPA process times derived from the more normally distributed square root values. The mean value for the lowest region, "Central," was either 4.5 years or 4.2 years, depending on which base of calculation was used.
• Geographical differences in the length of time to complete the NEPA process can also be considered on the basis of "Coastal" and "Central" regions, with the longest times taken in the "Coastal" region during the 1995 — 2001 study period. The mean value for the "Coastal" region was 5.5 years, based on actual NEPA process times, or 5.1 years, based on equivalent NEPA process times derived from the more normally distributed square root values. The mean value for the "Central" region was either 4.5 years or 4.2 years, depending on which base of calculation was used.
• A significant but weak correlation was found between the number of state agencies commenting on the EIS and the length of time to complete the NEPA process.
• A significant but weak correlation was found between the EPA rating indicating its level of concern for the EIS and the length of time to complete the NEPA process.
• A significant and sizeable difference in the length of time to complete the NEPA process was found between those projects completed during the 1995 — 2001 study period where economic / fiscal impacts were controversial issues and those where they were not.
• A significant and sizeable difference in the length of time to complete the NEPA process was found between those projects completed during the 1995 — 2001 study period where water quality impacts were controversial issues and those where they were not.
• A significant and sizeable difference in the length of time to complete the NEPA process was found between those projects completed during the 1995 — 2001 study period where more than one public hearing was held and those where no more than one public hearing was held.
• Although not found to be a statistically significant finding, it was determined that the requirement for a Coast Guard Bridge permit may extend the length of the NEPA process by six months to more than seven months, depending on which set of mean values is used.
• Although not found to be a statistically significant finding, it was determined that the requirement for a Section 4(f) evaluation may extend the length of the NEPA process by less than five months to as many as seven months, depending on which set of mean values is used.
• Although not found to be a statistically significant finding, it was determined that the requirement for a NPDES permit may extend the length of the NEPA process by between three and four months, regardless of which set of mean values is used.
• Two factors found to have an influence on the length of the NEPA process during the Phase I study (i.e., whether or not a Section 404 permit was required and whether or not noise was an issue) were not found to be significant nor related factors in the 1995 — 2001 Phase II database.

4.2 Conclusions

The purpose of this Phase II NEPA baseline study was, at least in part, to ascertain if the results found in Phase I would be repeated, or if a comparative assessment could be made between the two sets of results, thereby further identifying the baseline history of the length of time to complete the NEPA EIS process. As it turned out, it was possible as a result of this most recent study to get an even better understanding of the time involved in completing the process, at least in terms of establishing a time trend. As discussed in Section 3.1 and summarized in Section 4.1 above, it can be concluded from this current research that the length of time involved in fully complying with the NEPA process has continued to get longer in recent years in comparison to earlier decades. It can also be concluded that, whereas there previously existed a substantial variation in the length of time to fully comply with NEPA in different parts of the country, the various regions seem to have become much more uniform in recent years in terms of NEPA time requirements. In fact, it appears that the regional effect on the length of the NEPA process has become much broader than may have been the case previously, with the differences occurring primarily between the coastal states and the middle portion of the country. Given these findings, this study achieved the basic goal of expanding upon the baseline condition, and in identifying continuing trends.

As a result of these basic findings, it can also be concluded that the need for ongoing and future environmental streamlining efforts has never been greater. The fact that it has recently taken longer, on average, to prepare and complete an EIS than ever before, and that six former FHWA regions recently exhibited higher EIS completion times than did the highest region in the earlier Phase I study period, may be indicative of the fact that there is substantial opportunity for improvement in the overall process. The findings of this study also suggest that any improvement in the process could potentially result in a time reduction that would, at best, only achieve a total NEPA process time that was representative of an earlier decade. Nevertheless, any improvement and resulting time reduction would be a step in the right direction.

The one goal that was not successfully achieved by this study was the identification and confirmation of factors and conditions that may have a direct or indirect impact upon the NEPA process. Although the NEPA process seems to vary by broad geographic region, it does not seem to vary with the majority of the other variables tested in this data set. Those factors that were found to have a statistically significant relationship with the length of time to complete the NEPA process were either weak in their relationship, or difficult to explain the relationship in a practical sense. Other factors were not found to have a significant statistical relationship with the length of time to complete NEPA, even though some of them actually appear to have an influence on the length of the process. The factors identified in this study as being related to the length of time to complete NEPA were also different from those factors identified during the Phase I study, the reason for which is not entirely evident. In fact, the only factor that was found to have a noticeable effect on NEPA process time in both phases of study was the presence or absence of a Section 4(f) evaluation. However, even in that case, the factor was not found to have a statistically significant relationship with NEPA time in the current study, although, as in Phase I, a substantial time difference between projects involving 4(f) and those not involving 4(f) was found.

It can, perhaps, be further concluded from the lack of statistical variation in the length of the NEPA process when considered in relation to other project and process­related factors, that the process may have become more closely affected by external social, economic and attitude factors associated with broad geographic regions of the country. If this theory is true, then it would account for the inability to explain with any certainty, the causal relationships of NEPA time with the specific factors tested in this study. Although such a theory may make it difficult to take better control of the process by refining or varying some of the specific project or process-related factors, it would at least be understood that environmental streamlining efforts should also consider more global conditions that could affect the process in a given location. If nothing else, it reinforces the overall complexity of the NEPA process and the broad zone of influence that external conditions may have on its application.

In final conclusion, it is anticipated that the overall findings of this study should make a useful contribution to the further understanding of the length of the NEPA process, and in the refinement of a baseline for future research in evaluating the performance of environmental streamlining.

Back to Top

5.0 LIST OF PREPARERS

U.S. Department of Transportation, Federal Highway Administration

Kreig Larson
M.U.R.P., Urban and Regional Planning, University of Southern California, 1975
B.S., Soil and Water Science, University of California — Davis, 1972
Contribution: FHWA Project Coordinator

The Louis Berger Group, Inc.

Kenneth J. Hess, A.I.C.P., P.P.
M.C.R.P., City and Regional Planning, Rutgers University, 1977
B.A., Geography, University of Delaware, 1974
Contribution: Project Director

Antony B. Mason, Ph.D.
Ph.D., Geography, Rutgers University, 1989
M.A., Geography, Rutgers University, 1983
B.A., History, Swarthmore College, 1955
Contribution: Statistical Analysis

Wade Souders
B.A., Communications, University of North Texas, 1999
Contribution: Data Collection

Natalia Valderrama
M.A. candidate, Geography, City University of New York Hunter College, 2003 (expected)
B.A., Political Science, University of California, 1992.
Contribution: Data Collection

Matt Sumpter
M.S., Transportation Technology and Policy, University of California, Davis, 2000
B.S., Environmental Studies and Planning California State University-Sonoma, 1997
Contribution: Data Collection

Robert S. Hutchinson
M.S., Environmental Policy, NJIT, 1998
M.I.P., Infrastructure Planning, NJIT, 1997
B.S., STS, Environmental Studies, NJIT, 1995
Contribution: Data Collection Study Design