Skip to contentUnited States Department of Transportation - Federal Highway Administration FHWA HomeFeedback

Design-Build E ffectiveness Study

As Required by TEA-21 Section 1307(f)
Final Report
Prepared for:
USDOT - Federal Highway Administration
January 2006


< Previous (SEP-14 Innovative Contracting)Table of ContentsNext (Conclusions & Recommendations) >

IV. FINDINGS

This chapter presents the results of the various fact-finding efforts performed during this study to address the issues and concerns posed by Congress in Section 1307(f) of TEA-21 regarding the application of design-build contracting to projects in the Federal-aid highway program. The study findings are organized into the following eight sections that include the areas of inquiry posed by Congress:

The impacts of delivery approach on project duration, cost, and quality, as reported by the respondents to the project surveys, are established by using several approaches whose combined results provide a profile of the prevalent effects of design-build versus design-bid-build project delivery. These approaches include the following:

The comparison of actual results for similar groupings of design-build and design-bid-build projects provides the most objective basis for determining the relative impacts of using each project delivery approach on project duration, cost, and quality. The small sample size for these direct comparisons limits the statistical representation of the results. However, by comparing the combined results for each group of projects, organized by project delivery approach, the results are more representative than would be provided by individual one-on-one comparisons. This is due to each project's inherent uniqueness and the potential for that uniqueness to skew the results in some way, thereby making the comparison less useful for the purpose of this study.

OVERVIEW OF SEP-14 DESIGN-BUILD PROGRAM

Out of the 29 STAs, toll agencies, and local public agencies (collectively referred to as agencies) that responded to the design-build program survey, 18 agencies provided information about the size of their overall design-build program compared to their total program, broken down by type of project. Given the difficulty of assigning costs for multi-year projects to a particular year, the survey requested information on the number and costs for design-build projects completed in 2002 (the last full year of design-build project activity prior to the conduct of the surveys for this study), and for all projects completed by the agency in that year. These responses enabled the research team to assess the relative size of agency design-build programs when compared to the total agency program, by type of project and overall. The results of these responses are described below.

Extent of Design-Build Program

The responding agencies with design-build programs completed 73 design-build projects in calendar year 2002, representing a reported $1.2 billion in costs. This compared to 3,034 total projects completed that same year, at a total cost of $7.4 billion. Hence, while design-build projects represented only 2.8 percent of the total projects completed in 2002 for these combined agencies, they comprised 25.5 percent of the total costs for these projects. This is illustrated in Exhibits IV.1 and IV.2 in the "Total" column for each exhibit respectively.

Exhibit IV.1: Design-Build Projects as a Proportion of Total Projects
Completed in 2002 for Responding Agencies

Design-Build Projects as a Proportion of Total Projects

Source: D-B program survey: Q18, 13 responses

Exhibit IV.1 shows the proportion of projects completed in 2002 by responding agencies that were delivered using the design-build approach. The largest proportion of projects using design-build were for ITS, Road-New/Widen, and Bridge/Tunnel, ranging from 4 percent to 9 percent. In terms of number of projects, design-build remains a small percentage of the total programs in responding agencies at 2.8 percent.

Exhibit IV.2: Design-Build Project Costs as a Proportion of Total Project Costs
Completed in 2002 for Responding Agencies

Design-Build Project Costs as a Proportion of Total Project Costs

Source: D-B program survey: Q18, 13 responses

Exhibit IV.2 shows the proportion of overall design-build program costs for projects completed in 2002 by responding agencies. These percentages are much higher than in Exhibit IV.1 for all project types. While the proportion of project costs remains highest for the same project types as shown in Exhibit IV.1, the relative positions have shifted to favor more costly and perhaps risky project categories, such as Bridge/Tunnel and Road-New/Widen projects. These two project types show the highest increase in percentage when measuring the cost of projects versus the number of projects completed in 2002.

The higher percentage of total project costs using design-build indicates that in most cases, design-build projects tended to be significantly larger, more complicated projects. This is confirmed by Exhibit IV.3, which shows that the average cost of design-build projects completed in 2002 by agencies responding to the design-build program survey was $27.7 million, more than eleven times the average cost of all other projects completed that year (just under $2.4 million).

Exhibit IV.3: Average Project Costs by Delivery Method
Design-Build versus Non-Design-Build

Average Project Costs by Delivery Method

Source: D-B program survey: Q18, 13 responses

Contracting Methods Used by States with Design-Build Programs

The most frequently used contracting method reported by agencies in the design-build program survey is traditional design-bid-build, either by itself or combined with some kind of warranty (either material and workmanship or performance). As shown in Exhibit IV.4, 87 percent of the value of agency highway programs is reported to be delivered using some form of design-bid-build. Second in popularity is design-build, either alone or in combination with a warranty, which represents just over 9 percent of the program value. Besides in-house (force account) work, there is no other project delivery method with any significant use by the reporting agencies.

The predominant use of design-bid-build contracting applies across all types of projects, as shown in Exhibit IV.5. Design-build is most prevalent for Road-New/Widen, Rehabilitation/Reconstruct, and Bridge/Tunnel project types, ranging from 8 percent to 11 percent of the projects. The use of design-bid-build contracting ranges from 83- to 95-percent. This is a higher percentage than reported in Exhibit IV.1, which reflects only projects completed in 2002.

Design-build program managers responding to the program surveys estimated only about 2 percent of the design-build projects involved some other innovative contracting approach. In contrast, 21 percent of the design-build projects reported by design-build project managers involved another innovative contracting approach.

Exhibit IV.4: Contracting Methods Used for Agency Programs

Contracting Methods Used for Agency Programs

Source: D-B program survey: Q19, 21 responses

Exhibit IV.5: Contracting Methods Used by Project Type

PROJECT TYPEIN - HOUSE (FORCE ACCOUNT)DESIGN - BID - BUILD CONTRACTDESIGN - BID - BUILD WARRANTY CONTRACTDESIGN - BUILD CONTRACT
New/Widen2.6%83.1%3.2%11.2%
Rehab/Reconstruct2.9%84.3%3.9%8.9%
Resurface4.7%84.6%3.5%7.2%
Bridge/Tunnel2.5%85.8%3.2%8.4%
ITS0.0%94.5%0.0%5.5%
All Project Types3.0%84.4%3.4%9.3%

Note: The Design-Build Contract column includes both Design-Build and Design-Build Warranty contracts.

Source: D-B program survey: Q19, 21 responses.

The distribution of innovative contracting approaches for the full sample of design-build projects surveyed was as follows:

For the subset of design-build projects that were compared to similar design-bid-build projects, the distribution of innovative contracting approaches was as follows:

Procurement Methods Used for Design-Build Projects

For design-build projects completed in 2002, the predominant procurement method was Low Bid, as shown in Exhibit IV.6 at 56 percent. Best Value was used for 38 percent of the design-build projects. The only other procurement methods indicated were Alternative Bids/Designs and Multi-Parameter Bidding, each representing about 2 percent. This is in marked contrast to design-bid-build projects that were overwhelmingly low-bid based.

Exhibit IV.6: Procurement Methods Used for Design-Build Projects

Procurement Methods Used for Design-Build Projects

Source: D-B program survey: Q20, 14 responses

As shown in Exhibit IV.7, this distribution is representative of most of the project types, including Road New/Widen, Rehabilitate/Reconstruct, and Resurface. ITS projects show an even split between Low Bid and Best-Value.

Exhibit IV.7: Procurement Methods Used for Design-Build Projects by Project Type

PROJECT TYPELOW BIDALTERNATIVE BIDS/DESIGNSMULTIPARAMETER BIDDINGBEST-VALUE
New/Widen54.2%0.0%2.9%42.9%
Rehab/Reconstruct64.9%0.0%1.7%33.3%
Resurface59.0%0.0%1.0%40.0%
Bridge/Tunnel56.6%8.3%1.7%33.4%
ITS48.6%0.0%1.3%50.1%
All Project Types56.0%3.9%2.2%37.9%

Source: D-B program survey: Q20, 14 responses

Payment Methods Used by States with Design-Build Programs

The preferred payment method for design-build projects is lump sum, as shown in Exhibit IV.8. The only other payment method noted is Unit Price at 13 percent. This payment preference occurs for all project types, as shown in Exhibit IV.9, and reflects the transfer of project risk to the design-builder who is held responsible for satisfactory project completion and paid on that basis. The small portion of the design-build project costs falling under unit pricing is primarily due to the use of a combination approach to payment methods, whereby certain items are paid for on a unit price basis, while the majority of items are included in the lump sum (fixed price).

Exhibit IV.8: Payment Methods Used for Design-Build Projects

Payment Methods Used for Design-Build Projects

Source: D-B program survey: Q21, 16 responses

Exhibit IV.9: Payment Methods Used for Design-Build Projects by Project Type

PROJECT TYPEUNIT PRICELUMP SUM
New/Widen17%83%
Rehab/Reconstruct11%89%
Resurface13%88%
Bridge/Tunnel10%90%
ITS14%86%
All Project Types13%87%

Source: D-B program survey: Q21, 16 responses

Suitability of Design-Build Project Delivery

Projects of many sizes and complexities have used design-build project delivery over the years since the inception of the SEP-14 Innovative Contracting program, as shown in Exhibit III.3 in the last chapter. However, as Exhibit III.4 revealed, the overwhelming majority of SEP-14 program costs have been for projects over $100 million in cost. This reflects the perceptions of design-build program managers surveyed for this study. Based on the results of the program survey, design-build program managers rated the following project types as most suitable for design-build project delivery, as shown in Exhibit IV.10:

Least suitable among the project types is Road-Resurface. The suitability rating for design-build contracting is highly correlated to the size of the project, wherein the suitability rating more than doubles when going from small projects to mega projects. When deciding which projects to apply design-build contracting, medium to large projects (over $10 million) are considered the most suited to this project delivery approach.

When asked to rate the degree to which design-build project delivery would likely be used in the future, managers of design-build programs indicated a modest level of use (average rating of 2.9 out of a possible 6.0) for all project types, as shown in Exhibit IV.11. The most likely project types to use design-build in the future were Road-New/Widen and Rehabilitate/Reconstruct. The least likely project type for design-build was perceived to be Road-Resurface. This is indeed reflective of the actual usage statistics for the SEP-14 Design-Build Program provided in Chapter III.

Exhibit IV.10: Perceived Suitability of Design-Build Project Delivery
(Scale: 1 - Not Suitable; 6 - Highly Suitable)

Perceived Suitability of Design-Build Project Delivery

Source: D-B program survey: Q22, 26 responses

Exhibit IV.11: Likely Future Use of Design-Build Project Delivery
(Scale: 1 - None; 6 - Significant)

Likely Future Use of Design-Build Project Delivery

Source: D-B program survey: Q23, 27 responses

Legislative and Policy Requirements

Seventy-two percent of the agencies responding to the design-build program survey needed special permission or legislation to use design-build contracting. Half of the agencies with design-build programs required special legislation. Other requirements or restrictions included:

Fifty-nine percent of the agencies have developed written policies to guide their design-build programs. Seventy-one percent of these agencies had these policies in place before the first design-build project was initiated. The level of involvement by transportation agencies in formulating these policies ranged from none to very significant, with most having a modest to moderate level of involvement (averaging 3.6 on a 6-point scale). These results reveal the significant statutory and regulatory impediments that had to be overcome to enable agencies to apply SEP-14.

Adequacy of Design-Build Procurement and Contract Administration Procedures

Respondents to the design-build program survey generally characterized their agency's design-build procurement and contract administration procedures as adequate (averaging 4.7 on a 6-point scale for both). They also rated the resource levels available for procuring and administering design-build contracts as adequate (averaging 4.5 on a 6-point scale for both).

Among the contract administration procedures used for design-build projects, all of the respondents to the design-build program survey verified compliance with contract requirements as the project progressed to completion and/or upon project completion. Several of these agencies also had contract compliance verifications performed after project completion due to the inclusion of warranties. These verification procedures were specified in the design-build contract documents of the responding agencies.

EFFECTS OF DESIGN-BUILD CONTRACTING ON PROJECT DURATION, COST, AND QUALITY

Three of the key issues raised by Congress in requesting this study of design-build contracting in the Federal-aid highway program were the relative impacts of project delivery on project duration, cost, and quality. As part of this study, design-build project managers in STAs, toll agencies, and local transportation agencies involved in the SEP-14 program were surveyed to obtain available data and their perspectives concerning these three dimensions for specific design-build projects they were responsible for. As part of the project survey instrument, the respondents were asked to provide their best estimate of the percentage change in project duration, cost, and quality as a result of the design-build approach. The results of their responses are summarized in Exhibit IV.12 below. While this data reflects the recollections of design-build project managers and therefore may not be based on actual project records, it provides useful insight into how design-build project managers view their experience with a specific design-build project delivery.

Exhibit IV.12: Summary of Estimated Impacts of Using Design-Build on Project Duration, Cost, and Quality

Duration DimensionValueCost DimensionValueQuality DimensionValue
Responses62Responses48Responses61
Average-14.1%Average-2.6%Average0.0%
Median-10.0%Median0.0%Median0.0%
Mode-0.1%Mode0.0%Mode0.0%
Maximum50.0%Maximum65.0%Maximum10.0%
Minimum-63.0%Minimum-61.8%Minimum-10.0%
StandardDeviation24.4%Standard Deviation20.5%Standard Deviation2.1%

Source: D-B project survey: Q18, 48-62 responses

Exhibit IV.12 reveals that on average, the design-build projects had a distinctly greater potential for schedule reduction than cost reduction, as estimated by design-build project managers in their completed design-build project surveys. The average reduction in project duration was 14 percent, while the average reduction in project cost was almost 3 percent. There was no appreciable difference in project quality associated with project delivery approach, with most indicating no change. For both duration and cost impacts, there was a wide range of impacts, both positive and negative, that is reflected in the high standard deviations in these two data sets. This suggests that many other factors besides delivery approach impact the duration and cost of projects.

The following sections discuss in greater detail the impacts of design-build versus design-bid-build project delivery on the duration, cost, and quality of highway infrastructure projects developed under the SEP-14 program. The first section addresses the issue of project duration.

EFFECTS OF DESIGN-BUILD CONTRACTING ON PROJECT DURATION

The impacts of project delivery approach on project duration and the potential for project duration to change during the development process are presented in several ways in this section. This variety of information reflects the different ways in which survey participants responded to questions concerning the duration of design-build projects by project phase and relative to similar design-bid-build projects. Some of the information is based on estimates provided by survey respondents, while other information is based on actual data from sampled design-build projects and, when provided, from similar design-bid-build projects. When taken together, these various results provide a profile of schedule impacts that is indicative of the influence that the choice of project delivery approach can have on project length, both total and by phase.

The effect of project delivery on project duration can be determined in a number of ways. One way is to ask managers of actual design-build projects to estimate the impact of design-build project delivery on overall project duration. Another way is to compare the differences between planned and actual duration of project phases as the project moves from RFP development to completion. The effects of design-build project delivery on overall project and phase duration were developed in this study based on the combined results for the 69 completed project surveys, using respondent estimates and actual project schedule data.

A third method of measuring the impact of project delivery approach on project duration is to use cross-sectional data to compare the relative changes in project duration during the phases of project development between similar design-build and design-bid-build projects. This can be done by comparing either individual results for two very similar projects or the average results for a group of similar pairs of projects. Reported project-specific schedule changes by phase varied widely between different pairs of similar design-build and design-bid-build projects. To avoid the problem of the inherent differences between individual projects distorting the reported results, the analysis was based on comparing the average results by project delivery approach for the paired reported projects.

Out of the 17 pairs of projects reported, 11 pairs had sufficient data reported in the completed surveys to enable changes in project duration by phase to be developed. It should be noted that this is a relatively small sample that may not be statistically representative of the SEP-14 Program of projects or design-bid-build projects. Hence care needs to be taken in developing or applying any conclusions that are based on the results from this sample of paired design-build and design-bid-build projects.

Estimated Impacts of Design-Build on Project Duration

The project survey results revealed that design-build project delivery, in comparison to design-bid-build, had a mixed impact on project duration depending on the project type, complexity, and size. As Exhibit IV.13 shows, the estimated impacts of project delivery on project schedule resulted in a wide range of schedule variations, ranging from a 63-percent reduction to a 50-percent increase. This is reflected in the high standard deviation for this sample of estimates.

Exhibit IV.13: Estimated Change in Project Duration due to Design-Build Project Delivery

Duration DimensionValue
Responses62
Average-14.1%
Median-10.0%
Mode-0.1%
Maximum50.0%
Minimum-63.0%
Standard Deviation24.4%

Source: D-B project survey: Q18

When considered as a group, the surveyed design-build project managers estimated an average decrease of 14 percent in delivery time relative to design-bid-build. Out of 62 responses, 45 estimates were for schedule reductions and only 7 estimates indicated a schedule increase. Overall, 89 percent of the design-build project managers estimated no increase in project duration due to the application of design-build. These results suggest that from the perspectives of design-build project managers, project delivery approach (i.e., design-build versus design-bid-build) can be a significant factor in controlling and expediting project delivery schedules.

The range and average differences in procurement and contract administration time between design-build and design-bid-build project delivery approaches, as estimated by program survey respondents, is illustrated in Exhibit IV.14. Program survey respondents estimated that the time required for procurement of design-build contracts versus design-bid-build contracts ranged from 45-percent less to over 100-percent more time, with an average increased procurement time of 15 percent for design-build contracts. About two-thirds of the program survey respondents believed design-build projects had a longer procurement time than design-bid-build projects. In contrast, the actual time required for the administration of design-build contracts ranged from 75 percent less to 55 percent more time, with an average decrease in contract administration time of 3 percent for design-build contracts.

Exhibit IV.14: Range and Average Differences in Procurement and Contract Administration Time for Similar Design-Build and Design-Bid Build Projects

ActivityAverageMaximumMinimum
Contract Administration-2.8%55.0%-75.0%
Procurement15.0%105.0%-45.0%

Source: D-B program survey: Q12, 27 responses

These results suggest that design-build projects are perceived to take more time to set up and procure, but once awarded, require slightly less time for the contracting agency to administer in comparison to similar design-bid-build projects. The wide variation in responses reflects the newness of design-build procurement and contract administration processes and the diversity of project types and sizes for which design-build project delivery is used by transportation agencies.

Planned versus Actual Project Duration

Another consideration is how the duration of design-build projects changed from what was planned to what actually occurred. To provide the same basis for comparing project duration between design-build and design-bid-build project, total project duration is defined in this section as the time from advertising the design-build project (following preparation of the RFP), to completing the project as signified by contracting agency acceptance. Construction phase duration is defined as the time from initiating construction activity to acceptance of the project by the contracting agency.

Relative to what was planned before the surveyed projects began, total project duration declined by 0.9 percent on average while construction duration increased by 1.0 percent. Exhibit IV.15 shows a wide range of differences between planned and actual delivery times for the surveyed design-build projects. The same number of projects experienced a decrease in duration (15) as experienced an increase in duration (15) for the total project and construction phase. Four of the reported design-build projects did not experience any change in total project or construction phase duration.

Exhibit IV.15: Range and Average Differences in Planned versus Actual Total Project and Construction Phase Duration for Design-Build Projects

Project PhaseAverageMaximumMinimumStandard Deviation
D-B Construction Phase1.0%67.5%-54.7%28.5%
D-B Total Project-0.9%31.9%-35.5%15.4%

Source: D-B project survey: Q15, 51 responses for construction phase and 34 responses for total project

Design-Build versus Design-Bid-Build Project Duration

Another indication of the effect of design-build project delivery on project duration is obtained from the subset of 11 comparable design-build and design-bid-build project surveys completed for this study. For the purposes of this analysis, the time associated with preparation of procurement documents prior to advertising for bid is excluded from the definition of total project duration in Exhibit IV.16. This includes:

The pre-advertisement phase is subject to numerous influences beyond the control of contracting agency or respondents and not related to the choice of project delivery method.

Exhibit IV.16: Average Percent Change in Planned Versus Actual Total Project and Construction Phase Durations For Similar Design-Build and Design-Bid-Build Projects

Project PhaseAverageMaximumMinimumStandard Deviation
D-B Construction Phase-1.2%30.6%-54.7%27.3%
D-B-B Construction Phase11.6%71.7%-27.2%28.7%
D-B Total Project-4.2%23.1%-42.5%20.8%
D-B-B Total Project4.8%30.6%-20.9%14.9%

Source: similar D-B and D-B-B project surveys: Q15, 11 responses per survey type

As shown in Exhibit IV.16, on average the design-build projects achieved shorter total project duration and construction duration than originally planned. In contrast, the similar design-bid-build projects incurred longer timeframes, on average, for both total project and construction phase durations than originally planned. The subset of design-build projects reduced the planned project duration by an average of 4 percent, while the comparable design-bid-build projects increased total project duration by an average of 5 percent. This represents a 9-percentage point differential in total project duration between similar sets of design-build and design-bid-build projects.

Meanwhile, the subset of design-build projects had a decreased construction duration averaging 1 percent, while the comparable design-bid-build projects increased construction duration by an averaging 12 percent. This represents a 13-percentage point difference between actual and planned project timeframes.

The results of the program and project surveys, including both project manager estimates and actual project documentation, supports the claim that the design-build approach can reduce the overall duration of a project, in certain cases significantly. Despite wide variations in changes to project duration among the surveyed design-build and design-bid-build projects, particularly for the construction phase, the results revealed that longer than planned contract development and evaluation timeframes and potentially longer construction timeframes could be more than offset by certain features of the design-build process. These features included:

Exhibit IV.17 illustrates the general sequence of project development activities for both design-build and design-bid-build contracts. The two schedules demonstrate how the type of project delivery approach may influence the sequencing and duration of standard highway project development phases. The key feature that distinguishes these two project delivery approaches is the placement of design functions relative to the construction functions and the potential for overlap between the design and construction phases for the design-build approach.

These factors resulted in shorter total project durations than originally planned on average for the surveyed design-build projects, whereas these same timeframes increased for the surveyed design-bid-build projects. Interestingly, for the sampled design-build and design-bid-build pairs, the average planned project duration (excluding procurement document preparation) was longer for the design-build projects as a group, but actual project duration was shorter.

Exhibit IV.17: Sequence of Project Delivery Activities by Contract Approach

This schematic diagram shows the sequence of project delivery activities for both the traditional design-bid-build method and the design-build method. The overlap in the final design and construction phases for design-build provides the potential for shortened project delivery times.

Source: Dr. Keith Molenaar, University of Colorado at Boulder

The ability for design-build contractors to have greater control to better integrate the design and construction functions and to use parallel processing of certain functions previously required by contract and regulation to be done sequentially provide significant opportunities for trimming the time it takes to deliver a design-build project in comparison to its design-bid-build counterpart. Numerous respondents to the project surveys noted the ability to expedite a needed project as the primary motivation for using the design-build approach to project delivery.

EFFECTS OF DESIGN-BUILD CONTRACTING ON PROJECT COSTS

As with the previous section on project duration, the impacts of project delivery approach on project cost and the potential for project cost to change during the development process are presented in several ways in this section. This variety of information reflects the different ways in which survey participants responded to questions concerning the costs of specific design-build projects by project phase and relative to similar design-bid-build projects. Some of the information is based on estimates provided by project survey respondents and some of the information is based on actual cost data provided for sampled design-build projects and, when provided, similar design-bid-build projects. When taken together, these various results provide a profile of cost impacts that is indicative of the various impacts that the choice of project delivery approach can have on project costs, both total and by phase.

The effect of project delivery on project costs can be measured in a number of ways. One method is to use time series data to compare the level of project cost as the project moves from budget to contract to completion. Three measures of project cost change are developed in this section based on the phase of the project development process:

Average project cost changes by phase were developed in this study based on the combined results for the 69 completed project surveys, using actual project cost data. This provided the largest sample to determine these cost change impacts for design-build projects in the study.

As with project duration, another method of measuring the impact of project delivery approach on project cost is to compare the relative changes in project cost during the phases of project development between similar design-build and design-bid-build- projects. This can be done by comparing either individual results for two very similar projects or the average results for a group of similar pairs of projects. Project-specific cost changes by phase vary widely between different pairs of similar design-build and design-bid-build projects. To avoid the problem of the inherent differences between individual projects overly distorting the reported results, the analysis is based on comparing the average results by project delivery approach for the group of paired projects.

Out of the 17 pairs of projects reported, 11 pairs had sufficient data reported in the completed surveys to enable changes in project costs by phase to be developed. It should be noted that this is a relatively small sample that may not be statistically representative of the SEP-14 Program of projects or design-bid-build projects. Hence care needs to be taken in developing or applying any conclusions that are based on the results from this sample of paired design-build and design-bid-build projects.

Estimated Impacts of Design-Build on Project Cost

The project survey results revealed that design-build project delivery, in comparison to design-bid-build, had a mixed impact on project cost depending on the project type, complexity, and size. As Exhibit IV.18 shows, the estimated impacts of project delivery on project cost were wide-ranging, extending from a 62-percent reduction to a 65-percent increase. This is reflected in the high standard deviation for this sample of estimates.

Exhibit IV.18: Estimated Change in Project Cost due to Design-Build Project Delivery

Cost DimensionValue
Responses48
Average-2.6%
Median0.0%
Mode0.0%
Maximum65.0%
Minimum-61.8%
Standard Deviation20.5%

Source: D-B project survey: Q18

When considered as a group, the surveyed design-build project managers estimated an average decrease of 2.6 percent project cost relative to design-bid-build. Out of 48 responses, 20 estimates were for cost reductions, 17 for no change, and 11 for a cost increase. Overall, 77 percent of the design-build project managers estimated no increase in project cost due to design-build. These results suggest that from the perspectives of design-build project managers, project delivery approach (i.e., design-build versus design-bid-build) can be a contributing factor in controlling and potentially reducing project costs. However, project delivery approach is perceived to be less of a factor in affecting project cost than other characteristics of the project or its participants.

Reported Impacts of Design-Build on Project Cost

When actual project cost information is used from the project surveys, the design-build projects on average experienced no appreciable change in total cost (ranging from a decrease of 42 percent to an increase of 63 percent). As shown in Exhibit IV-19, this resulted from an average cost decrease of 2.3 percent between concept budget and contract, and an average cost increase of 3.2 percent during the contract phase.

When considering individual project results, the zero percent average total project cost change was the result of off-setting cost increases and cost decreases that both varied widely. Slightly more of the surveyed design-build projects experienced a decline in total project cost from budget to completion than experienced an increase, although the cost increases tended to be a somewhat higher percentage than the cost decreases per project. One-half of the design-build projects experienced a change in total cost within plus or minus 10 percent. One-quarter of the design-build projects experienced a decline in total cost of 10 percent or more and one-quarter experienced an increase of 10 percent or more.

Exhibit IV.19: Reported Change in Project Costs for Surveyed Design-Build Projects

DimensionAward Growth ((Award-Budget) / Budget)Contract Growth ((Final-Award) / Award)Total Growth ((Final- Budget) / Budget)
Responses363636
Average-2.3%3.2%0.0%
Median-0.1%0.5%-0.9%
Mode0.0%0.0%N/A
Maximum63%37%63%
Minimum-45%-42%-42%
Standard Deviation21%12%22%

Source: D-B project survey: Q16

In considering the average 3.2 percent increase in cost during the contract phase, most of the increase was due to design-builder increases of 5 to 7 percent, as shown in Exhibit IV-20.

Exhibit IV.20: Reported Changes in Design-Build Project Costs after Contract Execution

Reported Changes in Design-Build Project Costs after Contract Execution

Source: D-B project survey: Q16, 36 responses

The major compensating factors were decreases in the administrative costs to the contracting agency for preliminary engineering, contract preparation, contract administration, and quality control inspections, all of which represent significantly smaller portions of the project costs.

Comparison of Reported Project Cost Change Between Design-Build and Design-Bid-Build Project Delivery

The design-bid-build projects demonstrated somewhat more favorable cost results than their design-build project counterparts, as shown in Exhibit IV.21. As indicated by the results, there is a wide range of project costs for each project phase that is reflected by the large standard deviations shown for both design-build and similar design-bid-build survey samples. This suggests that many other factors besides project delivery approach are influencing the results.

Exhibit IV.21: Supporting Data for Reported Changes in Project Costs for Similar Design-Build and Design-Bid-Build Projects

Design-Build Projects

DimensionAward Growth ((Award-Budget) / Budget)Contract Growth ((Final-Award) / Award)Total Growth ((Final- Budget)/Budget)
Responses111111
Average1.9%6.0%7.4%
Median2.4%1.6%2.4%
ModeN/AN/AN/A
Maximum23%21%40%
Minimum-41%-4%-28%
Standard Deviation17%9%17%

Design-Bid-Build Projects

DimensionAward Growth ((Award-Budget) / Budget)Contract Growth ((Final-Award) / Award)Total Growth ((Final- Budget)/Budget)
Responses999
Average-1.4%4.3%3.6%
Median-0.9%0.4%-3.9%
ModeN/AN/AN/A
Maximum27%29%64%
Minimum-18%-3%-13%
Standard Deviation15%10%24%

Source: similar D-B and D-B-B project surveys: Q16

In considering the increases in cost during the contract phase of both the subset of design-build projects and similar design-bid-build projects, most of the increases occurred during the construction phase of the projects, as shown in Exhibit IV.22. For this small sample of similar projects, there was less cost growth indicated for the design-bid-build projects.

Exhibit IV.22: Comparison of Actual Reported Changes in Project Costs after Contract Execution for Similar Design-Build and Design-Bid-Build Projects

Comparison of Actual Reported Changes in Project Costs after Contract Execution for Similar Design-Build and Design-Bid-Build Projects

Source: similar D-B and D-B-B project surveys: Q16, 9-11 responses per survey type

Causes of Project Cost Changes

As noted earlier, project costs tend to experience most cost growth from contract award to project completion. Respondents to the design-build project survey were asked to rate the reasons for major changes in project costs (using a six-point scale ranging from "No Impact" (0) to "Major Impact" (6). This resulted in a series of ratings for the various factors listed in Exhibit IV.26, many of which are outside the control of the design-builder.

According to Exhibit IV.23, the leading cause of project cost changes was change orders: Owner required additions or subtractions had an average rating of 3.2, followed by design-builder or contractor suggested additions or subtractions at 2.4. The relative impacts of other factors on the cost of design-build projects are also shown below, with most between 1.6 and 2.0.

When the subset of design-build projects is compared to similar design-bid-build projects, as shown in Exhibit IV.24, the results reveal similar ratings of the causes of project cost changes for the two project delivery approaches. This suggests that both project delivery approaches are sensitive to similar factors, with design-build projects being significantly more sensitive to delays, additions, or subtractions caused by third parties than design-bid-build.

Exhibit IV.23: Causes of Cost Changes for Design-Build Projects

Causes of Cost Changes for Design-Build Projects

Source: D-B project surveys: Q16, 64 responses

Exhibit IV.24: Causes of Cost Changes for Similar Design-Build and Design-Bid-Build Projects

Causes of Cost Changes for Similar Design-Build and Design-Bid-Build Projects

Source: similar D-B and D-B-B project surveys: Q16, 17 responses per survey type

Change Orders and Claims

As noted above, a significant factor affecting project cost, as well as project duration and quality, is change orders. Exhibit IV.25 lists the number and effect of change orders, as well as claims, on the costs of design-build projects surveyed during the study.

Exhibit IV.25: Reported Change Order and Claim Activity for Design-Build Projects

Change Order DimensionChange Orders Per ProjectChange Order Costs Per Project ($000)Cost Per Change Order ($000)Claims Per ProjectClaims Cost Per Project ($000)
Responses6161616262
Average18$3,722$1220.3$80
Median8$400$290$0
Mode0$0$00$0
Maximum187$73,000$1,1695$3,000
Minimum0$0$00$0
Standard Deviation30$12,813$2371$429

Source: D-B project survey: Q16

Change orders represented 4.7 percent of the total costs for the surveyed projects. The average total value of change orders per project was $3.7 million. On average, reported design-build projects experienced 18 change orders per project. Out of 61 design-build projects reported, seven had more than 40 change orders. Only 12 projects had change orders with a total value of greater than $2 million. The average design-build change order equaled $122,000. The large standard deviation shown in Exhibit IV.26 reflects the wide range of change order experience among the surveyed projects.

Claims represented less than one-tenth of one percent of total project costs, with an average value per surveyed design-build project of $80,000 for claims. While there were few reported claims per design-build project, the average reported design-build project claim was $225,000. Claims affected less than 10 percent of the design-build projects reported.

Exhibit IV.26 lists the number and effect of change orders and claims on the costs of the subset of comparable design-build and design-bid-build projects surveyed during the study.

Exhibit IV.26: Change Order and Claim Activity for Similar Design-Build and Design-Bid-Build Projects

Design-Build Projects

Change Order DimensionChange Orders Per ProjectChange Order Costs Per Project ($000)Cost Per Change Order ($000)Claims Per ProjectClaims Cost Per Project ($000)
Responses1616161817
Average 16$837$850$0
Median14$467$350$0
Mode17$400N/A0$0
Maximum49$3,355$4726$0
Minimum4$14$10$0
Standard Deviation013$890$1191$0

Design-Bid-Build Projects

Change Order DimensionChange Orders Per ProjectChange Order Costs Per Project ($000)Cost Per Change Order ($000)Claims Per ProjectClaims Cost Per Project ($000)
Responses1414131818
Average22$588$470.6$337
Median8$275$470$0
Mode5N/A$500$0
Maximum80$4,000$1804$6,000
Minimum0$0$30$0
Standard Deviation27$1,013$491$1,413

Source: similar D-B and D-B-B project surveys, Q16

As shown in Exhibit IV.26, the subset of design-build projects had fewer change orders than the comparable design-bid-build projects, but the average cost per change order was greater for the design-build projects. This could be attributed to the greater size of design-build projects. This can be confirmed by the fact that change orders represented about the same share of total project costs for both design-build and design-bid-build projects. In contrast, the dollar value of claims per project was significantly lower for design-build projects than for comparable design-bid-build projects, with the subset of design-build projects having no reported cost of claims.

The various levels of cost change (growth and decline) indicated in this section for design-build projects from the program and project surveys reflects the difficulty in isolating the cause of cost changes and the influence of project delivery approach on cost control. The effect of project delivery approach on project costs is difficult to determine due to the many other factors beyond the control of the contract team than can influence the final project cost. Since both design-build and design-bid-build projects experienced a wide range of cost changes during development, project costs appeared to be more influenced by factors independent of project delivery approach. Design-build project delivery appeared to reduce agency costs of contract administration and inspection relative to design-bid-build project delivery. Of particular note was the reduced level of claims and their related impacts on cost growth for the full survey sample of design-build projects reported, and especially for the subset of design-build projects when compared to similar design-bid-build projects.

EFFECT OF DESIGN-BUILD CONTRACTING ON PROJECT QUALITY

Contracting agencies are interested in obtaining a quality project, as well as one that is completed in a timely and cost-effective manner. Quality can be defined in a number of ways, depending on the point of view of the evaluator and the aspect of the project being considered. For many, project quality is defined as meeting all project specifications and their prescribed standards. As engineered projects, conformance with project specifications is determined by testing project materials and inspecting the end product relative to these standards. Project acceptance is based on the results of these tests and inspections.

Project quality can also be measured by determining if the contracting agency is satisfied with the product. Contracting agency satisfaction can have many dimensions, ranging from knowing the project meets all specifications (input-based measure of quality) to being pleased with the performance of product (outcome-based measure of quality). Performance-based specifications focus on the results or outcomes of the project and can be measured by various criteria, such as ride quality, durability, and visual aesthetics. These can be either quantitative or qualitative criteria and are subject to the expectations of the contracting agency.

Prescribed (standards-based) specifications are the traditional way of determining project acceptance. However, with more responsibility being given to contractors for delivery of highway projects, the use of performance (outcome-based) specifications is increasing as a way to account for project quality dimensions not captured by standards and specifications and to promote greater innovation by contractors to achieve more cost-effective projects of equal of better quality.

In this section, project quality is discussed in terms of the following three criteria:

Estimated Impacts of Design-Build on Project Quality

The design-build project survey responses indicated that application of design-build project delivery had no differential impact on project quality in the opinion of the survey respondent. According to Exhibit IV.27, most (93 percent) of the design-build projects performed at the same level of quality as those delivered by the design-bid-build approach. Three percent of the surveyed projects note an increase in project quality while the same small percentage noted a decrease in project quality. Of the projects that experienced an increase in project quality, the average improvement was 8.5 percent, while the average decrease in project quality for projects that experienced a decline was 7.5 percent. These results are based on estimates provided by project survey respondents regarding changes in project quality by applying the design-build project delivery approach.

Exhibit IV.27: Estimated Change in Project Quality due to Design-Build Project Delivery

Quality DimensionValue
Responses61
Average 0.0%
Median0.0%
Mode0.0%
Maximum10.0%
Minimum-10.0%
Standard Deviation2.1%

Source: D-B project survey: Q. 18

Contracting Agency Satisfaction

Contracting agency satisfaction with the outcome and process of project delivery is one of the primary ways to measure the quality of different approaches. According to project survey respondents, 97 percent of the design-build projects have fulfilled their intended purpose. Half of the respondents indicate that the method of project delivery has had a significant impact on the outcome of the project.

As shown in Exhibit IV.28, project survey respondents express a high level of satisfaction with design-build projects, averaging 4.7 on a six-point scale (in which 1 is poor and 6 is superior). The same high level of contracting agency satisfaction is noted in the compliance with warranties and standards & specifications.

Exhibit IV.28: Contracting Agency Satisfaction Ratings of Design-Build Projects
(Scale: 1 - Poor; 6 - Superior)

Contracting Agency Satisfaction Ratings of Design-Build Projects

Source: D-B project survey: Q17, 69 responses. Out of the 69 surveyed projects, 26 had warranty provisions.

Based on a detailed statistical analysis of project survey responses, the research team discovered that overall contracting agency satisfaction is highly correlated with the following project characteristics:

The results of this analysis are summarized in Exhibit IV.29. These results are statistically significant at the 95 percent confidence level.

Exhibit IV.29: Overall Contracting Agency Satisfaction by Project and Contract Type

Project/Contract CharacteristicOverall Agency Satisfaction
LowerHigher
Procurement MethodLow BidBest Value
Project TypeRoad-Resurface/RenewalRoad-New/Widen and Rehabilitate/Reconstruct
Project SizeSmallerLarger
% of Design Completed at AwardHigherLower

Source: D-B project survey: Q2, 4, 10, and 17, 69 responses

When a subset of 19 design-build projects is compared to similar design-bid-build projects, the survey results indicate that overall contracting agency satisfaction with design-build projects is on a par with design-bid-build projects, as shown in Exhibit IV.30. However, conformance with warranty provisions and standards and specifications are both rated higher for design-build projects than for similar design-bid-build projects.

Exhibit IV.30: Comparison of Contracting Agency Satisfaction Ratings between Similar Design-Build Projects and Design-Bid-Build Projects
(Scale: 1 - Poor; 6 - Superior)

Comparison of Contracting Agency Satisfaction Ratings between Similar Design-Build Projects and Design-Bid-Build Projects

Source: similar D-B and D-B-B project surveys: Q17, 17 responses per survey type

Experience of the Project Delivery Team

Another measure of the quality of project delivery is the relative experience of members of the project team for the various functions comprising the project. Those teams with highly qualified and experienced members are likely to perform the best in delivering a quality project consistent with the terms of the contract. When asked to characterize the prior experience and expertise of key design-build project stakeholders, respondents to the design-build project surveys provided the responses listed in Exhibit IV.31.

As shown in Exhibit IV.31, design-build project survey respondents perceived that insurance and finance officials were the most experienced parties, with a 3.8 rating on a six-point scale (in which 1 equals no experience and 6 equals significant experience). Designers, builders/contractors, design-builders, and agency-sponsors were next most experienced, with ratings ranging from 3.0 to 3.3. Subcontractors and subconsultants were rated the least experienced, with ratings of 2.1 and 2.4 respectively. These modest ratings reflected the relative newness of the design-build project delivery approach to the domestic highway industry.

Exhibit IV.31: Perceived Experience among Stakeholder Groups for Design-Build Projects
(Scale: 1 - None; 6 - Significant Experience)

Perceived Experience among Stakeholder Groups for Design-Build Projects

Source: D-B project surveys: Q17, 69 responses

When responses to the subset of design-build project surveys were compared to those for similar design-bid-build projects, respondents perceived stakeholders in the design-bid-build process had a much greater level of expertise and familiarity. These results, shown in Exhibit IV.32, are consistent with the prevalence of design-bid-build contracting by transportation agencies and the relative newness of design-build contracts.

It is interesting to note the high average experience rating given to the Agency/Owner category for design-bid-build projects (5.8) and the much lower experience rating given to this same category for design-build projects by the survey respondents (who themselves are part of this stakeholder group). This further highlights one of the challenges facing those interested in pursuing design-build project delivery - raising the expertise and experience in this approach among contracting agencies and thereby increasing their comfort in applying design-build at a more significant level.

Exhibit IV.32: Perceived Experience among Stakeholder Groups for Similar Design-Build and Design-Bid-Build Projects

Perceived Experience among Stakeholder Groups for Similar Design-Build and Design-Bid-Build Projects

Source: similar D-B and D-B-B project surveys: Q17, 17 responses per survey type

Other Project Delivery Success Criteria

Survey respondents report a number of factors they use to measure project delivery success. Chief among them is meeting the objective quality standards of the contracting agency, plus project completion on time and under budget. These and other project delivery success criteria are outlined in Exhibit IV.33. Each of these factors relates in some way to the issues of concern posed by Congress in requesting this study.

Exhibit IV.33: Project Delivery Success Criteria Used by Project Survey Respondents

Quality

  • Project quality relative to comparable design-bid-build projects
  • Number of claims or change orders
  • Achievement of project scope and objectives, including project quality standards, traffic impacts, and environmental goals

Cost

  • Total project cost relative to budget
  • Amount of cost overrun
  • Cost of claims or change orders

Timeliness

  • Project opening relative to scheduled completion date
  • Length of project extension
  • Project advancement or velocity relative to schedule

Other

  • Dollar amount of incentive payment to contractor relative to maximum possible incentive payment
  • Ability to control cost and schedule to issue toll road revenue bonds at minimum risk
  • Success of implementing new technology or construction techniques
  • Experience of contractor with design-build projects or other projects similar in scope to the design-build project
  • Project likelihood without use of design-build or other approaches to advance project
  • Implementation of extended warranty or other risk mitigation approaches

Source: D-B project survey: Q17, 48 responses

APPROPRIATE LEVEL OF PRELIMINARY DESIGN FOR DESIGN-BUILD PROCUREMENTS

Having chosen design-build contracting to deliver a particular project, contracting agencies must decide at which point in the project development life cycle to initiate the design-build contract. This decision is influenced by the nature and complexity of the project, the needs of prospective design-build teams to understand the full requirements and potential risks of the proposed project before developing and offering a design-build contract proposal, the comfort level that contracting agencies have in letting design-builders develop the scope of the project based on the project's defined performance objectives, and what has become accepted practice based on other, earlier design-build projects.

An earlier survey of six STAs using design-build found a broad range for the level of preliminary design completed before issuing requests for bids of proposals for design-build projects.1 The range was 15 percent to 50 percent, with the average among the six agencies being 31 percent (Colorado and Washington, respectively). The higher the percentage preliminary design completed before design-build procurement the more likely the selection process was based on low bid (New Jersey and Indiana). The lower the percentage preliminary design completed the more likely the selection process was based on a composite score or best-value (South Carolina and Arizona).

Based on the completed design-build project surveys, Exhibit IV.34 shows the distribution of the percentage completion of preliminary design relative to other pre-construction activities such as right-of-way acquisition, permit acquisition, and environmental review.

Exhibit IV.34: Percent Completion of Selected Functions at Design-Build Project Award

Percent Completion of Selected Functions at Design-Build Project Award

Source: D-B project surveys: Q10, 69 responses

A high proportion of right-of-way acquisition, permit acquisition, and environmental review functions are completed by design-build contract award, while most preliminary designs are below 30 percent complete by design-build award. Notice that several projects had the design-builder responsible for all of these functions (the projects with functions at 0 percent completion by design-build award, to the far left of the chart).

On average, as shown in Exhibit IV.35, right-of-way acquisition was 89 percent complete for surveyed design-build projects, permit acquisition was 83 percent complete, and environmental clearance2 was 99 percent complete.

Exhibit IV.35: Average Percent Completion of Selected Functions at Design-Build Project Award

Average Percent Completion of Selected Functions at Design-Build Project Award

Source: D-B project surveys: Q10, 69 responses

Permit acquisition may include Section 404, navigable waterways, water quality, air quality, noise levels, and other local permits. NEPA clearance may include an environmental assessment or preparation of a full environmental impact statement. Most of the surveyed design-build projects had completed 100 percent of right-of-way acquisition and permit and environmental clearance prior to project award. On average, design was 27 percent complete prior to design-build contract award. For 81 percent of the reported projects, the percentage of design completion by design-build contract award was 30 percent or less.

When a subset of design-build projects is compared to similar design-bid-build projects, survey results reveal that design-build projects had somewhat less right-of-way acquisition and permit and environmental clearance complete by contract award, as shown in Exhibit IV.36. All comparable design-bid-build projects had completed 100 percent of preliminary project activities. For the subset of design-build projects surveyed, the average percent design completion prior to going to a design-build contract was 37 percent, with 78 percent of the projects at 30 percent or less.

Exhibit IV.36: Average Percent Completion of Selected Functions at Project Award for Similar Design-Build and Design-Bid-Build Projects

Average Percent Completion of Selected Functions at Project Award for Similar Design-Build and Design-Bid-Build Projects

Source: similar D-B and D-B-B project surveys: Q10, 17 responses per survey type

These results are consistent with the finding in Exhibit IV.29 that the level of contracting agency satisfaction reported for design-build projects was higher for lower levels of preliminary design completed before design-build contract award. This could be attributed to a design-builder's ability to influence the project design earlier in the process to promote its constructability and cost-effectiveness. While each project should be considered on an individual basis, the results suggest that no more than 30 percent of preliminary design be completed before design-build contract award, with lower percentages as the contracting agency gains more experience with design-build contracting and greater reliance is placed on performance-based specifications.

IMPACT OF DESIGN-BUILD CONTRACTING ON SMALL BUSINESSES

The advent of design-build project delivery has raised concerns by some that small firms3 may be unable to participate on design-build teams, particularly as the design-build team lead or prime contractor, due to the increased functional scope and scale of many design-build contracts, more stringent qualification requirements, and/or higher bonding requirements. In some cases, contracting agencies have applied design-build to smaller projects to address this and other issues. In the context of this report, small business participation includes the involvement of smaller firms in design-build projects as a prime contractor, joint venture partner, or subcontractor.

Agency respondents to the design-build program survey indicated that the percentage of design-build project costs going to small businesses was about the same on average as design-bid-build projects, with only a very small reduction indicated for design-build projects. This is shown in Exhibit IV.37, in which small business involvement with design-build projects is estimated to be within 2 percent of the level of involvement with design-bid-build projects overall, and within 1 percent of that for design-bid-build projects when the design-build team is based locally. These results suggest that small businesses are not disadvantaged when projects are developed through the design-build process, according to agency design-build program managers.

Exhibit IV.37: Small Business Involvement on Similar Design-Build and Design-Bid-Build Projects

Competitive DimensionDesign-BuildDesign-Bid-Build
AverageMaximumMinimumAverageMaximumMinimum
Percent of Project Costs Provided by Small Firms31.3%55.0%5.0%33.0%55.0%15.0%
Percent of Project Costs Provided by Small Firms on Local Competing Teams32.3%75.0%5.0%32.9%75.0%15.0%

Source: D-B program survey: Q15, 15 to 22 responses

Size of Prime Contractors and Subcontractors for Design-Build Versus Design-Bid-Build Projects

Two-thirds of agency design-build program respondents indicated that on average, the prime contractors and subcontractors for design-build projects are similar in size to their counterparts on design-bid-build projects. The remaining one-third indicated that prime contractors for design-build projects were significantly larger than their counterparts for design-bid-build projects (5.4 on a 6-point scale with 1 being smaller and 6 being larger), while subcontractors for design-build projects were only marginally larger in size than their counterparts for design-bid-build projects (3.4 on the same 6-point scale).

While the size of prime contractor firms may have been somewhat larger for design-build projects than for design-bid-build projects (though not always so), the size of subcontractor firms was essentially the same. To the extent small businesses are currently involved in the design and construction of design-bid-build projects, similar opportunity appears to exist for design-build projects, particularly in the role of subcontractor. These results suggest small businesses are playing a comparable role for design-build projects as for design-bid-build projects, and that the design-build project delivery process is not preventing small businesses from participating in design-build projects to a comparable degree.

Limits on the Extent of Design-Build Contract Value Held by the Prime Contractor

The opportunity for small businesses to participate in design-build projects is also affected by the amount of the contract retained by the prime contractor. Where maximum limits are defined, the contracting agency determines the extent to which firms other than the prime contractor must be involved in the project. Where minimum limits are defined, the contracting agency determines the limits to which firms other than the prime contractor can be involved in the project. The larger the contract value and the higher the percentage of contract value required to be retained by the prime contractor both suggest fewer opportunities for involvement by small businesses that are less likely to have the resources or background to warrant serving as the prime contractor.

On a program-wide basis, 81 percent of the respondents indicate there are maximum limits and/or minimum limits on prime contractor involvement specified in design-build contract language. Where maximum limits existed, the maximum percentage ranged from 70 percent to 100 percent. Where minimum limits existed, the minimum percentage ranged from 30 percent to 51 percent. FHWA eliminated the 30 percent self-performance requirements for traditional contracts when it developed the Design-Build Contracting Regulations4.

Use of Direct Hire Versus Subcontractor Resources for Design-Build Contracts

The project survey results indicate that for design-build projects, an average of 60 percent of design work was subcontracted, with the remaining 40 percent handled as direct hire (self-performance by the design-builder or its core team members). As shown in Exhibit IV.38, an average of 75 percent of construction work was directly hired and 25 percent was subcontracted.

Exhibit IV.38: Proportion of Direct Hire and Subcontracted Work by Function for Design-Build Projects

Analysis DimensionDirect Hire DesignSubcontracted DesignDirect Hire ConstructionSubcontracted Construction
Responses48484848
Average40%60%75%25%
Median4%96%85%16%
Mode0%100%100%0%
Maximum100%100%100%100%
Minimum0%0%0%0%
Standard Deviation45%45%32%32%

Source: D-B project survey: Q13

Based on the smaller sample of similar design-build and design-bid-build projects shown below in Exhibit IV.39, the project survey indicates that design-bid projects had a much higher percentage of subcontracted design work than similar design-bid-build projects, averaging 52 percent for design-build projects versus only 11 percent for design-bid-build projects. In contrast, the proportion of subcontracted construction work was about the same for design-build as for design-bid-build projects, at 21 percent to 24 percent, respectively. This may be due to the predominant role of construction contractors on many design-build teams, who may be more willing to subcontract design work than construction work. This may also be due to the larger size and complexity of many design-build projects, which require more sophisticated designs.

Exhibit IV.39: Proportion of Direct Hire and Subcontracted Work by Function for Similar Design-Build and Design-Bid-Build Projects

Design-Build Projects

Analysis DimensionDirect Hire DesignSubcontracted DesignDirect Hire ConstructionSubcontracted Construction
Responses11111111
Average48%52%79%21%
Median70%30%80%20%
Mode0%100%100%0%
Maximum100%100%100%45%
Minimum0%0%55%0%
Standard Deviation47%47%17%17%

Design-Bid-Build Projects

Analysis DimensionDirect Hire DesignSubcontracted DesignDirect Hire ConstructionSubcontracted Construction
Responses551111
Average89%11%76%24%
Median89%11%70%30%
Mode89%11%100%0%
Maximum100%20%100%42%
Minimum80%0%58%0%
Standard Deviation7%7%16%16%

Source: similar D-B and D-B-B project surveys: Q13. A smaller number of D-B-B project surveys reported a breakdown in design work between direct hire and subcontract resources.

These results suggest that design-build contracts may spread more of the design work among subconsultants than comparable design-bid-build contracts, which should be a positive feature for small business enterprises.

Prequalification Requirements

Another factor impacting the extent of competition for design-build projects is the extent to which proposers must be prequalified, which means having satisfied certain performance or capability criteria to be able to bid on design-build project contracts. All respondents to the design-build program survey indicated they require some form of prequalification, as noted in Exhibit IV.40. About half the respondents used a two-step project-specific process, whereby the initial step used prequalification information to select a short list of prospective proposers for design-build projects. The rest used either a one-step project-specific prequalification or a more general or annual prequalification to define eligible prospective proposers.

Exhibit IV.40: Prequalification Requirements for Design-Build Projects

Prequalification Requirements for Design-Build Projects

Source: D-B program survey: Q3, 29 responses

Extent of Competition for Design-Build Projects Versus Design-Bid-Build Projects

As reported in the design-build program surveys, the extent of competition for design-build projects is perceived to be significantly lower than that reported for design-bid-build projects. As shown in Exhibit IV.41, almost 40-percent fewer teams responded to requests for qualifications (RFQs) for design-build projects than to requests for pre-qualifications for design-bid-build projects; however, it is recognized that many states use an RFQ process for the design phase and an annual program-wide prequalification process for the construction phase of a design-bid-build project. One-third fewer teams responded to requests for proposals (RFPs) for design-build projects than to invitations for bids (IFB) for design-bid-build projects. Similarly, there were 40-percent fewer local design-build teams than local design-bid-build teams that responded per project opportunity. The design-build program survey also revealed that the proportion of design-build teams led by local firms was estimated by respondents to be 81 percent, versus 91 percent for design-bid-build teams.

The ability to offer stipends to unsuccessful proposers of design-build projects recognizes the relatively high cost of preparing a design-build proposal when compared to a design proposal or construction bid. By offering a stipend to pre-qualified proposers, contracting agencies also seek to increase the number of capable proposers and thereby enhance competition for these types of procurements. The program survey indicated that just over half of the design-build program respondents paid stipends to unsuccessful teams proposing on a design-build project, with the average approximately $50,000 per team.

Exhibit IV.41: Level and Type of Competition for Similar Design-Build And Design-Bid-Build Projects

Competitive DimensionDesign-BuildCompetitive DimensionDesign-Bid-Build
AveMaxMinAveMaxMin
Average Number of Teams Responding to Request for Qualifications (RFQ) per Project6153Average Number of Teams Responding to Prequalification per Project10400
Average Number of Teams Responding to Request for Proposals (RFP) per Project462Average Number of Teams Responding to Invitation for Bid (IFB) per Project6120
Average Number of Local Teams (Led by Local Firms) per Project351Average Number of Local Teams (Led by Local Firms) per Project5102
Average Amount of Stipends Paid per Team per Project ($000s)$48.8$250.0$0.0Average Amount of Stipends Paid per Team per Project ($000s)$0.0$0.0$0.0

Source: D-B program survey: Q15, 24 responses

These results indicate that the number of firms or teams responding to a design-build project was estimated to be smaller than that for design-bid-build projects, particularly at the local level. This may reflect the newness and perceived risks (including the higher costs of proposal preparation) associated with this particular project delivery approach to the Federal-aid highway program and the traditional design and construction firms that have served this program over the years. It may also result from the two-step selection process frequently used for design-build projects whereby only qualified firms are short-listed by the agency. This "short-listing" process limits the level of competition for these projects to avoid having too many firms commit the large level of resources typically needed to generate a design-build proposal. It also limits the financial exposure of the contracting agency if a stipend is offered to all unsuccessful proposers. This is not considered a detriment to the design-build procurement process since competition among qualified firms is retained.

The larger scale and scope of a typical design-build project, the more extensive use of short-listing to procure design-build services, and the newness of this project delivery method makes it impossible to compare the number of proposing teams for a design-build project and a similar design-bid-build project. Of note is the relatively high proportion of local teams reported to be proposing on design-build projects by agency program managers, which would tend to refute claims that design-build project delivery heavily favors national firms over local firms.

ASSESSMENT OF SUBJECTIVITY USED IN DESIGN-BUILD CONTRACTING

The advent of design-build project delivery has also raised questions regarding the extent of subjectivity used in selecting a design-build contractor team and the effect this can have on project cost-effectiveness and the equity of the procurement process. The primary issues involve the use of non-cost factors in determining the successful bidder for a design-build project and the perceived objectivity of the selection criteria and process used. This is of particular concern when design-build contracts are assigned on the basis of expected best-value to the contracting agency, versus the more traditional low-bid selection process.

To properly address this issue, it is important to understand what factors prompt contracting agencies to use design-build in the first place and to what extent cost is an important factor in their considerations. These and other factors that affect the extent of competition for design-build contracts are explored in this section from both program-level and project-level perspectives.

Design-Build Project Designation Criteria

On a program-wide basis, there are a number of factors that affect the decision of whether or not to use design-build. Exhibit IV.42 shows the relative rankings in descending order of eight factors (using a 6-point scale where 0 is unimportant and 6 is extremely important). According to the design-build program survey responses, the most important factor was viewed as the urgency of the project.

Exhibit IV.42: Relative Importance of Factors Considered in Deciding Whether to Use Design-Build
(Scale: 1 - Unimportant; 6 - Extremely Important)

Relative Importance of Factors Considered in Deciding Whether to Use Design-Build

Source: D-B program survey: Q1, 29 responses

Among the design-build program survey respondents, 97 percent considered project urgency of great importance to making this decision. The next most important factor was opportunity for innovation, followed by opportunity for risk transfer. The other five factors, headed by project cost and quality, were not viewed as important as project urgency in deciding whether to use the design-build approach.

Respondents rated federal initiatives that encourage agencies to consider design-build project delivery, even on an experimental basis such as SEP-14, as having the least influence on their decisions regarding the application of this project delivery approach to particular projects among the factors noted above. Other factors considered important by individual respondents but not included in the average ratings shown above included project size, project type, and funding availability.

These results suggest that for early users of design-build in the Federal-aid highway program, the potential for faster project delivery and the application of innovative approaches served as the primary motivators for their decision to use design-build project delivery for projects. The importance of project delivery speed reinforces the earlier findings in this chapter that show that design-build offered the greatest potential for reducing project duration than for improving any other key project performance criteria.

Design-Build Contract Award Criteria

For those projects designated for design-build delivery, respondents to the design-build program survey perceived cost as the most important factor in awarding project contracts even though project duration was the most important factor in deciding whether to use the design-build approach. As shown in Exhibit IV.43, cost and cost combined with duration were perceived to be the most important factors in awarding design-build project contracts by the design-build program managers from responding agencies.

Exhibit IV.43: Key Factors Considered in Awarding Design-Build Project Contracts
(Scale: 1 - Unimportant; 6 - Extremely Important)

Key Factors Considered in Awarding Design-Build Project Contracts

Source: D-B program survey: Q2, 29 responses.

Overall, cost represented just over half of the weighting (55 percent) when factors other than cost were included in the proposal evaluation. In several cases, respondents noted that their traditional project award process and criteria had not changed for design-build projects - namely low bid-based project award (cost only - 100 percent weighting). Other approaches used included adjusted scoring based on the weighting of factors unique to the project and technical merit.

These results are reflective of the design-build project survey responses, where just over half of the design-build projects were procured on the basis on low-bid, versus 42 percent using best-value and 4 percent using multi-parameter approaches. This is illustrated in Exhibit IV.44.

Exhibit IV.44: Procurement Approach of Design-Build Projects

Procurement Approach of Design-Build Projects

Source: D-B project survey: Q4, 29 responses

Among the design-build projects procured based on best-value, a diversity of evaluation approaches was used, as illustrated in Exhibit IV.45 (see Appendix A - Glossary of Terms for definitions of each best value method). Forty-one percent of the reported projects were awarded based on an adjusted bid and 14 percent were awarded based on an adjusted score for a combined total of 55 percent. These two best value evaluation methods are quite similar (they are the mathematical reciprocal of each other) and both weigh price at 50 percent. Weighted criteria represent 28 percent of the best-value procurements. Cost-technical tradeoff, fixed price/best design, and low bid meeting technical criteria round out the remaining approaches.

These results suggest that while project urgency and innovation were the primary motivators for using design-build contracting, cost remains the primary factor for awarding design-build contracts, even when other factors such as duration, team reputation, and quality were included in the deliberations. In addition, low bid continued to play an important role in contract award decisions, with best-value approaches using multiple criteria including cost gaining momentum.

Exhibit IV.45: Best Value Procurement Evaluation Methods for Design-Build Projects

Best Value Procurement Evaluation Methods for Design-Build Projects

Source: D-B program survey: Q4, 29 responses

Since design-build includes a significant design element, it is important to include these other factors as is the case for the procurement of engineering service contracts (which must use qualifications-based selection procedures). Best value selection provides for the consideration of both cost and other more subjective factors such project management, quality control, and team reputation.

Given the results of this review of contracting approaches and features of design-build versus design-bid-build projects, it is apparent there remain significant controls in place to limit the extent of subjectivity in awarding design-build contracts and to preserve reasonable access for prospective contractors of all types and sizes to projects using this contracting approach.

OTHER DESIGN-BUILD CONTRACT FEATURES

There are a number of additional features of design-build contracts that can impact the relative risk to the public and private sector participants in the contract and the opportunity to apply more cost-effective approaches to accomplishing the objectives of the project. These include:

Each of these features is discussed below as they relate to the SEP-14 design-build projects surveyed in this study.

Design-Build Contract Pricing Approaches

The majority of the value of reported design-build projects used fixed price-lump sum pricing, as shown in Exhibit IV.46. This is in contrast to design-bid-build contracts that typically use unit pricing. One-quarter of the value of the reported design-build projects involved the use of unit pricing. As noted earlier, the smaller portion of design-build project costs falling under unit pricing or other approaches was primarily due to the use of a combination of payment methods, whereby certain breakout items were quoted on a unit price basis, while the majority of items were included in the fixed price-lump sum.

Exhibit IV.46: Contract Pricing Methods for Design-Build Projects

Contract Pricing Methods for Design-Build Projects

Source: D-B project survey: Q6, 69 responses

The use of fixed price-lump sum pricing by design-build contracts is a distinguishing feature that reflects greater project risks transferred to a design-build contract team. This form of contract pricing allows for progress billing and payment instead of detailed quantity measurement and verification. This simplifies and reduces the field administrative effort associated with contract billing and payment for the contracting agency and design-builder. Lump sum contract pricing can also be linked to performance standards, which can be used to trigger payments for work/service delivered. Lump sum pricing focuses attention on the project schedule and encourages the design-builder to stay within project scope, avoid change or extra work orders that are a major cause of project cost creep, and complete the project within the allotted timeframe.

Design-Build Contract Specifications

One of the purported advantages of design-build project delivery is the opportunity to use more performance-based specifications to encourage greater innovation by the design-build team and focus on project performance results versus conformance with product specifications that may be outdated given the latest technology and research. Some are concerned that the substitution of performance-based specifications may confer unfair advantage to those contract teams with access to proprietary technology that may satisfy performance standards more cost-effectively than their competitors.

Based on the results of the design-build project survey, the specifications used for over half of the reported design-build projects were reported to be entirely prescriptive. The remaining projects were reported to involve some combination of prescriptive and performance-based specifications. Only 3 percent of the responses were reported to use only performance-based specifications in their contract. About 10 percent of the projects had a 50/50 mix of performance and prescriptive specifications. When combined, the sample of design-build projects reflected a 73 percent use of prescriptive-based specifications and only a 20 percent use of performance-based specifications, as shown in Exhibit IV.47.

Exhibit IV.47: Average Relative Use of Prescriptive and Performance Specifications for Design-Build and Design-Bid-Build Projects

Survey SourceSample SizePrescriptive SpecificationsPerformance Specifications
D-B Survey Sample6973%20%
D-B Project Subset1758%34%
D-B-B Similar Projects1759%33%

Source: D-B project survey and similar D-B and D-B-B project surveys: Q11

In comparing a more limited sample of design-build projects to similar design-bid-build projects, the average relative use of prescriptive and performance specifications was similar for both project delivery methods. As shown in Exhibit IV.47, prescriptive specifications were used for nearly 60 percent of design-build and design-bid-build projects, while performance specifications were used for approximately 33 percent of design-build and design-bid-build projects in the sample. These results demonstrate the growing use of performance-based specifications for highway project contracts for both project delivery approaches.

Design-Build Contract Incentives and Disincentives

The use of incentives and disincentives in project contracts is intended to promote certain desirable project delivery results (such as early completion) and minimize undesirable consequences (such as unexcused completion delay or failure to meet specifications). Project incentives are becoming more popular for reconstruction and rehabilitation of existing roads to promote early project completion and thereby reduce inconvenience to motorists using the facility. Liquidated damage and disincentive provisions are frequently included in construction contracts to recover the contracting agency's construction engineering costs and road user costs resulting from contractor delays.

When questioned about the use of incentives and disincentives on completed design-build projects, 20 percent of the respondents to the project survey indicated the use of specific incentive clauses while 46 percent indicated the use of specific disincentive clauses. The various kinds of incentives noted in the design-build project survey responses are listed in Exhibit IV.48.

Exhibit IV.48: Design-Build Contract Incentive Types

Early Completion Incentive

  • Flat incentive for early completion
  • Daily incentive for early completion
  • Completion of specified elements such as roadway lighting and bridges

Traffic Management

  • Auxiliary lane availability
  • Travel time

Revenue Sharing

  • 70 percent of net toll revenue from early traffic

Project Quality

  • Pavement smoothness or ride quality
  • Materials quality
  • Workmanship quality

Other Incentives

  • Award fee for management, quality, and schedule
  • Project safety
  • Public relations program

Source: D-B project survey: Q7, 69 responses

Most frequently mentioned incentives were for early completion of the project or a specified element of a project, or for project quality including pavement smoothness. Other incentives were offered for traffic management, public information, project safety, and toll revenue sharing for early opening.

The various kinds of disincentives noted in the design-build project survey responses are listed in Exhibit IV.49. Disincentives included late completion penalties and stipulated damages as well as lane rental fees for the closure of traffic lanes and shoulders. The same relative usage and types of incentives and disincentives were found among the seven pairs of comparable design-build and design-bid-build projects. This suggests that the issue of incentive and disincentive use is more a function of evolving industry practice rather than project delivery approach.

Exhibit IV.49: Design-Build Contract Disincentive Types

Late Completion Penalties

  • Project schedule overruns
  • Escalated damages for extended delay
  • Failure to meet given materials and roadway smoothness standards

Stipulated Damages

  • Delay in substantial completion
  • Delay in final acceptance

Other Disincentives

  • Limited eligibility for time extensions
  • Lane rental fees for closing existing traffic lane and/or shoulder

Source: D-B project survey: Q7, 69 responses

Extended Warranties in Design-Build Contracts

Thirty percent of the surveyed design-build projects included extended warranties5. Of those design-build projects with warranties, two-thirds were material and workmanship warranties and one-third were performance or condition warranties. The duration of design-build project warranties ranged from six months to ten years. One ITS project included a two-year warranty for computer software. Most warranties were one, five, or seven years in duration, with the average duration of just over four years.

Most reported design-build project warranties included clauses that defined conditions that complete or void the warranty. These were typically defined in terms of time limits. Several projects had other "escape" clause criteria, including the impact of a natural disaster on the project or other factors beyond the contractor's control. Pavement warranties usually have an axle loading limit on the warranty to account for the impacts of vehicle use on pavement deterioration. However, none of the survey respondents indicated the use of an axle loading exclusion clause in their pavement warranty clauses.

In general, the inclusion of an extended warranty is not considered a competitive factor in the selection of design-build projects by the respondents. In addition, extended warranties are reported to have little or no impact on the quality, timeliness, and cost of design-build projects. Like incentives and disincentives, the use of extended warranties appears to be more a function of project type than project delivery approach. As a form of risk transfer to the project contracting team, they have been used for projects involving all kinds of project delivery approaches, not just design-build. This may reflect the fact that design-build contracts are used more for project expediency and innovation than to ensure the longevity of project performance.

Certainly extended warranties can be a distinguishing feature in promoting competitive products to consumers, as the automobile manufacturers have discovered. However, in the case of highway projects, the traditional separation of project development and preservation phases (operations and maintenance), both temporally and organizationally within agencies, likely diminishes the perceived value of extended warranties for those units responsible only for project development.

* * * * * * * * * *

Chapter V presents the conclusions and recommendations of the study, based on the findings presented in this chapter.


FOOTNOTES

  1. Molenaar, Keith R. and Douglas D. Gransberg, Design-Builder Selection for Small Highway Projects, ASCE Journal of Management in Engineering, Vol. 17, No. 4, October 2001
  2. Under the National Environmental Policy Act (NEPA) or similar state legislation
  3. Small business is defined as any organization with less than 500 employees and $6 million in average annual receipts for service organizations ($28.5 million for general building and heavy construction contractors and $12 million for special trade construction contractors) For applicable small business size standards by industry category, see the U.S. Small Business Administration's Small Business Size Regulations, 13 CFR §121 or the Table of Small Business Size Standards.
  4. 23 CFR §635.116(d)(1)
  5. Based on D-B project survey: Q8, 69 responses

< Previous (SEP-14 Innovative Contracting)Table of ContentsNext (Conclusions & Recommendations) >

FHWA Home | Feedback
FHWA