Indefinite Delivery/Indefinite Quantity Contracting Practices (2015)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

38 chapter four CONTRACT ADMINISTRATION PROCEDURES INTRODUCTION This chapter discusses the post-award administration proce- dures currently in use to successfully manage an IDIQ con- tract at the work order level. An agency’s outstanding ability to conduct pre-award activities is irrelevant if preconceived plans and strategies are not appropriately applied after award of the contract (Haukohl and Macicak 2012). In a tradition- ally procured contract, these plans and strategies are usually communicated to project managers through completed design, milestone schedules, budgets, and detailed scope of work for the entire contract. All of these elements constitute guidelines to ensure that the project proceeds as planned. However, to achieve the flexibility desired in an IDIQ contract, some or all of these elements are typically incomplete, making it harder to coordinate work between different project phases. Therefore, it is important to develop well-defined policies and procedures, such as those found in the study and presented in this chapter, to successfully accomplish work under an IDIQ contract. In addition to work order development and administration procedures, this chapter discusses three key issues found dur- ing the literature review and case example analysis that must be fully addressed in order to achieve a successful implementation of IDIQ techniques. These issues are payment for mobilization costs, escalation of construction pricing on multi-year IDIQ contracts, and the impact of IDIQ contracting on surety bonds. WORK ORDERS—DEVELOPMENT AND ADMINISTRATION Every project to be executed within an IDIQ contract begins with the issuance of work orders. In other words, a work order becomes the primary contract document and determines loca- tion, contract time, scope of work, and any other additional information required by the contractor to successfully per- form all work contained in the work order (MnDOT 2014). Figure 27 shows a generic work order development process synthesized from the information collected in this study. Specific DOT cases in the next chapter show more detailed procedures in accordance with the procurement policies and administration practices of each agency. Work orders become the primary contract document and deter- mine location, contract time, scope of work, and other project related information. To develop a work order, the owner first must identify a project and verify that it is suitable to be executed under a given IDIQ contract in accordance with the scope stated in the solicitation documents. If the project falls within the IDIQ contract scope, the owner proceeds to develop the work order to be performed by a contractor. Once the project is com- pleted by the contractor, the owner will verify that work and services provided meet the requirements stated in the work order. Upon approval of the work delivered, the owner pro- ceeds with final payment and work order close-out activities. If the owner rejects the work, the contractor will make correc- tions until it meets the acceptance criteria stated in the work order or contract documents. Work Order Scoping The scoping of work orders actually starts with the plan- ning of the contract. It is important that owners are careful when determining potential scope and expected size of work orders to be issued under the contract. IDIQ minimum guar- anteed amounts are typically established so that the agency is committed to award at least one work order to each awardee (Rueda 2013), and corresponds to effective practices in IDIQ contracts awarded by the CFLHD and FDOT (FDOT 2009; CFLHD 2012). Since the minimum value represents the worst case scenario for bidders, they may choose to prepare price pro posals based on this minimum amount of work. Therefore, work orders that are too small may encourage higher than normal bid pricing, and too large work orders may prevent the agency from reasonably awarding future work. A complete analysis of this issue allows the authors to propose an approach to establish an optimal scope and size of work orders. When planning these aspects of the contract, owners are to consider the average monetary size of poten- tial projects to be executed under the contract if traditional contracting methods are used. If the agency maintains this average size for all work orders, bids should be similar to those obtained using a single-project contracts (Rueda 2013). Contractor Involvement in Work Order Development Some authors and studies strongly recommend the use of interactive work order development procedures with high levels of communication between owners and contractors (OFPP 1997; Thornton 2002). Given that when using IDIQ techniques contractors are awarded the contract before all

39 the work orders are developed, at least after the initial one, agencies can request input during the preconstruction phase of work order development, in a manner similar to that used in CMGC contracts (Shane and Gransberg 2010). This phi- losophy was found in two state transportation agencies using single award IDIQ contracts; Missouri and New York State. The work order development procedures of these two agencies are explained in detail in chapter five. The two DOTs allow time during work order development to discuss project- related issues with contractors before issuing the final work order document as follows: • An initial meeting is held with the contractor to jointly discuss the scope of a given work order. • After the initial meeting, the agency prepares a draft of the scope, which is again reviewed by the contractor and returned to the owner with comments. • Once having developed the final work order document, it is send to the contractor who must prepare and submit a proposal. • Upon submission of this proposal, the parties have a final meeting to discuss any remaining issues before starting with the execution of the work order (NYSDOT 2009; MoDOT 2010). It is important to understand that this level of commu- nication and interaction between contractors and agencies is difficult to obtain in multiple award IDIQ contracts owing to the implications of receiving preconstruction input from multiple contractors. However, some state transportation agencies, such as MnDOT, have found a limited response from contractors when trying to involve them in work order development procedures for single award IDIQ contracts Work Order Placement Procedures— Multiple Award IDIQ As mentioned in chapter three, this study found different multiple award approaches used by different state DOTs. This section describes some different work order placement procedures used by one federal and four state agencies for multiple award construction/maintenance IDIQ contracts. Table 15 permits the reader to see some differences between the multiple award approaches implemented by these state transportation agencies. In addition, the inclusion of a federal agency in this table allows a comparison to be made between these two types of transportation agencies. Table 15 shows how some of the advertise/award methods discussed in chapter three can be also applied at the work order level. Although all state DOTs in Table 15 conduct non-competitive work order placement procedures, they are actually competing using bid packages and/or statements of qualifications originally submitted to compete for the contract. For instance, the California, Delaware, and Maine DOTs use FIGURE 27 Work order development process. Agency Number of Awardees Work Order Placement Procedure Caltrans Up to 6 Each work order is assigned to the contractor with the lowest total price for the specific bid line items required for that work order. Total price is calculated using unit prices originally submitted by the contractor. If this contractor is unable to perform the work, the work order is assigned to the awardee with the next lowest price. DelDOT All responsive and responsible bidders MaineDOT All responsive and responsible bidders VTrans All responsive and responsible bidders For this contract, Vermont DOT “will use discretion in selecting firms for each individual assignment. Assignments of work will be based on the cost to perform a particular assignment, the contractor’s availability and/or ability to complete an assignment within a specified time period” (VTrans 2011). CFLHD 3 Each work order is advertised to all awardees. Then, contractors submit price proposals used by CFLHD to select the low-bid. TABLE 15 WORK ORDER PLACEMENT PROCEDURES—MULTIPLE AWARD CONSTRUCTION/ MAINTENANCE IDIQ CONTRACTS

40 low-bid techniques to assign work orders using unit prices from contractors’ price proposals and pay items and quanti- ties determined for each project. A comparison between this table and Table 10 in chapter three shows that procurement methods used by these agencies to award contracts have been extended to the work order level. Payment Provisions Survey participants from state DOTs were asked to indicate the payment provisions used in their typical IDIQ contracts. Figure 28 shows their responses to this question. Although not as common as unit price contracts, fixed-price and cost- reimbursable contracts are also being used by some agencies. In addition, survey responses showed that 35% of the agencies that responded to this question are using a combination of these methods in a single contract. However, a review of IDIQ con- tract documents from 20 federal agencies showed a different trend with a greater use of fixed-price contracts (see Figure 29). The difference between these two approaches may be explained by Figure 30. Based on the observed use that public owners are given to payment provisions shown in Figures 28 and 29, a decision-making system was developed based on the complexity and uncertainty perceived in each project. The purpose of this decision-making system is to help agencies select suitable payment provisions for each project. When considering the use of the fixed-price projects shown in Figure 30, it can be inferred that the greater use of this payment FIGURE 28 Payment provisions—State DOTs. FIGURE 29 Payment provisions—Federal agencies. FIGURE 30 IDIQ payment provision decision system.

41 method by federal agencies is because these agencies handle larger budgets and execute larger and more complex projects. The information in Figure 30 also allows one to conclude that those contracts that permit a combination of payment provisions may have broad scopes for the issuance of differ- ent types of work orders. In other words, the use of multiple payment provisions in a single IDIQ contract allows agen- cies to undertake projects with different levels of complexity and uncertainty. The use of multiple payment provisions in a single IDIQ contract allows agencies to undertake projects with different levels of complexity and uncertainty. Construction Task Catalog The use of construction task catalogs in IDIQ contracts is a prac- tice adopted by at least seven agencies as indicated in the survey responses and literature review. The construction task catalog is intended to contain unit prices for all pay items required during the entire contract period. These unit prices include labor, equip- ment, and material costs (The Gordian Group 2011). Thus, contractors are required to bid a set of adjustment factors that represent profit and overhead. To calculate the total cost of each work order, the extended price of each item (unit price times work quantity) in the work order is multiplied by the respective adjustment factor. Usually interested contractors are required to bid a set of adjustment factors to be applied to each pay item in accordance with different working conditions. For instance, Table 16 shows two different sets of adjustment factors required by two different agencies; the New York State and Missouri DOTs. Chapter five and Appendix D present a detailed analysis of contracts executed by these agencies. Construction task catalogs (provided by owners) and adjustment factors (bid by contractors) are primarily used for two purposes; contractor selection and work order pric- ing procedures. Contracts from the New York State and Mis- souri DOTs analyzed in chapter five and Appendix D (whose adjustment factors are shown in Table 16) are both low-bid single award IDIQ contracts for construction/maintenance services. Unlike traditional low-bid IDIQ contracts, where the contract is awarded by totaling the extended price of all pay items, contractor selection procedures in these two con- tracts only total the extended prices of expected portions of work (in dollars) to be performed under the different working conditions associated with each adjustment factor. This pro- cess may be better understood with the following example, which was adapted from an example included in the IFB advertised by the Missouri DOT (MoDOT) (see Table 17). The example in Table 17 consists of calculating the total bid for a contractor that submitted the adjustment factor pre- sented in this table for a contract with a total base cost (before adjustment) equal to $100,000. In this example, estimated percentages of work performed during normal work hours, nighttime, and weekends are 85%, 10%, and 5%, respectively. The incorporation of adjustment factors into the selection formula supposes an additional source of risk. Taking into consideration the importance of maintaining a fair proportion- ality between bid quantities and actual quantities of work per- formed throughout the contract, the use of adjustment factors adds another element that should remain reasonably propor- tional during the project life cycle. In addition to maintaining a ADJUSTMENT FACTORS NYSDOT MoDOT Normal Work Adjustment Factor: 7:00 a.m. to 5:00 p.m. Monday–Friday Normal Work Adjustment Factor: 6:00 a.m. to 7:30 p.m. Monday–Friday Nighttime Work Adjustment Factor: 7:30 p.m. to 6:00 a.m. Monday–Thursday Other than Normal Work Adjustment Factor: 5:00 p.m. to 7:00 a.m. Monday–Friday All day Saturday, Sunday, and Holidays Weekend Work Adjustment Factor: 7:30 p.m. Friday–6:00 a.m. Monday Holidays TABLE 16 EXAMPLES OF ADJUSTMENT FACTORS Item Description Approximate Quantity AdjustmentFactors Bid Amount Normal Work Adjustment Factor 0.85 x $100,000 = $85,000 1.10 $85,000 x 1.10 = $93,500 Nighttime Work Adjustment Factor 0.1 x $100,000 = $10,000 1.15 $10,000 x 1.15 = $11,500 Weekend Work Adjustment Factor 0.05 x $100,000 = $5,000 1.22 $5,000 x 1.22 = $6,100 Total Estimated Cost $100,000 Bid Total $111,100 Adapted from MoDOT (2010). TABLE 17 MISSOURI DOT BID TOTAL CALCULATION EXAMPLE

42 sufficient level of uniformity on the use of contract pay items (as used to estimate approximate quantities of work in Table 17), agencies using this type of IDIQ contract would keep, to the maximum extent possible, a uniform distribution of work among the different working conditions, following the percent- ages used for the selection of the contractor. This observation is not intended to disqualify the use of construction task catalogs and adjustment factors with IDIQ contracts; instead, it under- lines this gap in knowledge as a potential topic for future study. During an interview with MoDOT, there was an opportu- nity to learn the specific reason that this agency adopted this approach. The first version of IDIQ contracts implemented by the agency used a more traditional approach in which bidders were required to submit unit prices for all pay items and bid quantities. However, MoDOT found that this approach was sus- ceptible to the submission of unbalanced bids by contractors. Therefore, this DOT decided to implement a bidding and pay- ment system based on a construction task catalog and adjust- ment factors to improve its budget control in IDIQ contracts. Mobilization This section discusses a major issue with regard to IDIQ contracting; compensation of contractors for mobilization expenses. This issue has been addressed in different ways by different agencies. For instance, the New York State and Mis- souri DOTs include mobilization pay items in their construction task catalogs to be used in accordance with the requirements of each work order. FDOT compensates contractors for mobiliza- tion expenses using a percentage of the construction cost. In the case of multiple award contracts with competitive work orders, mobilization expenses are included in the price for each work order. However, it is not included in the case of single award IDIQ contracts, where the contractor is only allowed to bid once when there is no precise information about exactly when, where, and how each work order will be performed. The team found four different approaches used to com- pensate contractors for mobilization expenses and asked DOT and contractor survey participants to rank them on a scale of 1 to 4 from the most suitable (1) to the least (4). They were also asked to indicate those alternatives that they consider not suitable at all. In addition, the repetition of rank position was allowed for those alternatives they considered equally valuable. This section analyzes responses from state DOTs, AGC/ARTBA contractors, and MnDOT contractors. Table 18 presents the responses from the three groups of IDIQ users. It appears to be clear that Option 3 is the most preferred alternative for most survey participants in all three State Departments of Transportation Option # Description 1 2 3 4 Not Suitable Total Responses Mean 1 Fixed percentage of the construction cost stated by the agency (owner) and applied to each Work Order. 4 1 3 6 8 22 3.59 2 Fixed percentage bid by contractors to be applied to each Work Order and factored into the selection of the low bid. 4 4 3 4 7 22 3.27 3 Fixed price bid by contractors to be used on each Work Order. 9 1 3 1 6 20 2.70 4 No separate mobilization pay item. Mobilization expenses are included in the bid items. 6 5 4 1 5 21 2.71 AGC/ARTBA Contractors Option # Description 1 2 3 4 Not Suitable Total Responses Mean 1 Fixed percentage of the construction cost stated by the agency (owner) and applied to each Work Order. 2 2 5 2 2 13 3.00 2 Fixed percentage bid by contractors to be applied to each Work Order and factored into the selection of the low bid. 4 5 3 0 1 13 2.15 3 Fixed price bid by contractors to be used on each Work Order. 7 3 2 0 1 13 1.85 4 No separate mobilization pay item. Mobilization expenses are included in the bid items. 1 1 1 6 5 14 3.93 MnDOT Contractors Option # Description 1 2 3 4 Not Suitable Total Responses Mean 1 Fixed percentage of the construction cost stated by the agency (owner) and applied to each Work Order. 4 7 10 8 8 37 3.24 2 Fixed percentage bid by contractors to be applied to each Work Order and factored into the selection of the low bid. 5 12 8 6 6 37 2.89 3 Fixed price bid by contractors to be used on each Work Order. 19 5 6 5 2 37 2.08 4 No separate mobilization pay item. Mobilization expenses are included in the bid items. 1 1 4 15 16 37 4.19 TABLE 18 MOBILIZATION COMPENSATION APPROACHES—SURVEY RESPONSES

43 surveys. However, the other positions in the ranking are not easy to identify. This drawback can be easily overcome by using the mean column to measure the acceptability of each option. This mean is calculated by assigning a value of 5 when a particular option was considered not suitable at all and using the respective value for the other responses (val- ues from 1 to 4). Thus, those alternatives with low overall rankings, with 1 being the minimum possible mean, will be preferred over those with high overall values, with 5 being the maximum possible mean. Figure 31 illustrates the difference in opinions between each group of participants regarding the appropriateness of each mobilization compensation option for single award IDIQ contracts. It presents the preferences for each group of survey participants in simpler form. This figure shows that all three surveys ranked Option 3, “fixed price on each work order,” as the most attractive alternative to compensate a contractor for mobilization expenses in a single award FIGURE 31 Mobilization compensation approaches—Suitability diagram.

44 IDIQ contract. This option consists of a fixed-price bid by contractors during procurement to be used on each work order. On the other hand, it can be noted that state DOTs consider the use of no separate mobilization pay items in the contract to be virtually as good as using a fixed price for each work order, an approach that was strongly rejected by both populations of contractors. A statistical analysis of data in Table 18 showed a strong correlation in the responses from both groups of contractors; therefore, it can be considered that both types of survey participants present the same opinion about all compensation alternatives listed in the table. Even more important, this analysis showed that the opinions of state DOTs and contractors with regard to the top ranked alter- native (Option 3) are significantly similar, confirming the preference for this mobilization compensation approach by most participants. A similar conclusion was obtained from the study con- ducted for MnDOT (Gransberg and Rueda 2014). However, the selection of the fixed price for each work order alter- native as the preferred approach was validated with input submitted by contractors and MnDOT’s staff. This input highlighted an important aspect to be considered; that dur- ing initial bidding there is high uncertainty regarding work order project locations and scopes. Therefore, it would be difficult for contractors to determine a rational “one-size- fits-all” mobilization price for every potential work order. This situation forces contractors to use the worst case sce- nario (i.e., mobilization to the most remote point in the contract area), and agencies would pay more for projects that do not match that scenario. Therefore, Option 3 was extended from one to multiple mobilization items to be bid by contractors. By asking contractors to bid on various mobilization pay items, anticipating different potential case scenarios, agen- cies have more flexibility to execute larger contracts cover- ing more locations with a single solicitation. Thus, contracts with broader scopes and with potential projects distributed in larger regions would require a greater number of mobiliza- tion pay items to counteract the scope and location uncer- tainty (Gransberg and Rueda 2014). Therefore, it can be concluded that a highly effective mobilization compensation approach would be one in which contractors are required to bid fixed prices on multiple mobilization pay items, whose applicability will be individually determined by the agency on a case-by-case basis in accordance with the scope and location of each work order project. By asking contractors to bid on various mobilization pay items anticipating different potential case scenarios, agencies would have more flexibility to execute larger contracts covering more locations with a single solicitation. Price Escalation Price escalation in multi-year IDIQ contracts was another major issue. This issue primarily applies to single award and multiple award contracts with non-competitive work orders. Given the absence of competition in the adjudication of work orders in these two cases, contractors are either required to maintain unit prices throughout the contract period (no esca- lation clauses) or expect a fair adjustment in contract prices in accordance with actual changes in the construction market. However, given the dynamics of the construction industry, the volatility of construction material prices, the uncertainty regarding actual construction activities, and the difficulty in determining a feasible distribution of work along the duration of an IDIQ contract, it is hard for contractors to accurately esti- mate unit prices for multi-year contracts, making it difficult for them to bid on long-term contracts with no escalation clauses (Rueda 2013). Conversely, if these escalation clauses are not consistent with actual changes in the construction market, con- tractors would be compelled to include larger contingencies in their price proposals. This study found that most contractors are willing to bid on multi-year IDIQ contracts even without esca- lation clauses, although it would represent higher construction costs for the agency. Following a similar procedure as the one presented in the previous section to find a suitable mobilization approach, participants in the contractors and DOT staff surveys were asked to rank seven different alternatives on a scale of 1 to 7, from the most suitable for IDIQ contracting (1) to the least (7). Also, participants were asked to indicate those alter- natives they do not consider suitable at all and repeat ranking positions in those cases that consider that two alternatives are equally valuable. The use of more alternatives makes Table 19 more difficult to read. However, again the mean column can be used to assess the preferences of each survey participant in each survey. Figure 32 illustrates the different opinions reported by each group of participants with regard to the appropriate price escalation approach for multi-year single award IDIQ contracts. In this case these surveys clearly show different opinions, unlike what happened with mobilization compensation approaches. An important observation from Figure 32 is that while for most state DOTs the use of no escalation clauses appears to be the most suitable alternative, this is the least preferred option for contractors. The similarity of opinions of both groups of contractors in relation to this price escalation alternative was confirmed by conducting a statistical test. Without considering Option 7, owing to the higher costs that it may represent for contractors (Rueda 2013) and the inconvenience that it appears to represent for contractors, Option 3 becomes the most suitable alternative for state DOTs and AGC/ARTBA contractors. It involves the use of existing national or local indexes such as the Producer Price Index published by the Bureau of Labor Statistics, the Construction

45 State Departments of Transportation Option # Description 1 2 3 4 5 6 7 Not Suitable Total Responses Mean 1 Fixed annual percentage stated by agency (owner) to adjust all bid items. 4 1 1 1 0 1 3 8 19 5.47 2 Fixed annual adjustment rate (%) bid by contractors to be applied to adjust all bid items factored into the selection of the low bid. 0 2 2 0 2 2 2 9 19 6.21 3 Using a national or local existing index (ENR, BLS, RSMeans). 3 3 2 1 2 1 3 4 19 4.63 4 Using a regional index (by district or similar) developed and kept by each agency using its historical bid data. 1 0 1 6 0 1 3 6 18 5.72 5 Using a state index by category developed and kept by each agency using its historical bid data (asphalt, concrete, structures, etc.). 1 0 3 3 0 2 4 6 19 5.79 6 Using an index by pay item. Measure the change of a pay item by using historical bid data from that or similar items. 1 3 0 1 1 3 3 7 19 5.84 7 No escalation. Bid prices are used along the base contract period and kept during potential contract extensions. 9 1 1 3 1 1 0 2 18 2.94 AGC/ARTBA Contractors Option # Description 1 2 3 4 5 6 7 Not Suitable Total Responses Mean 1 Fixed annual percentage stated by agency (owner) to adjust all bid items 2 3 1 1 3 1 1 1 13 3.92 2 Fixed annual adjustment rate (%) bid by contractors to be applied to adjust all bid items factored into the selection of the low bid. 4 0 2 2 3 1 1 1 14 3.86 3 Using a national or local existing index (ENR, BLS, RSMeans). 3 6 1 1 1 1 0 1 14 2.93 4 Using a regional index (by district or similar) developed and kept by each agency using its historical bid data. 3 0 5 1 2 2 0 0 13 3.38 5 Using a state index by category developed and kept by each agency using its historical bid data (asphalt, concrete, structures, etc.). 2 1 2 4 3 1 0 0 13 3.62 6 Using an index by pay item. Measure the change of a pay item by using historical bid data from that or similar items. 2 1 0 3 2 3 1 1 13 4.54 7 No escalation. Bid prices are used along the base contract period and kept during potential contract extensions. 0 0 0 2 1 0 5 6 14 6.86 MnDOT Contractors Option # Description 1 2 3 4 5 6 7 Not Suitable Total Responses Mean 1 Fixed annual percentage stated by agency (owner) to adjust all bid items 4 3 4 9 1 6 6 3 36 4.58 2 Fixed annual adjustment rate (%) bid by contractors to be applied to adjust all bid items factored into the selection of the low bid. 9 6 6 5 3 2 2 3 36 3.44 3 Using a national or local existing index (ENR, BLS, RSMeans). 2 9 5 5 2 1 6 6 36 4.47 4 Using a regional index (by district or similar) developed and kept by each agency using its historical bid data. 0 5 4 7 2 3 9 6 36 5.25 5 Using a state index by category developed and kept by each agency using its historical bid data (asphalt, concrete, structures, etc.). 0 2 2 11 6 2 6 7 36 5.39 6 Using an index by pay item. Measure the change of a pay item 4 3 4 7 3 5 5 5 36 4.72 by using historical bid data from that or similar items. 7 No escalation. Bid prices are used along the base contract period and kept during potential contract extensions. 1 3 0 2 3 3 9 15 36 6.42 TABLE 19 PRICE ESCALATION APPROACHES—SURVEY RESPONSES Cost Index (CCI) published by the Engineering New-Record, or the CCI issued by RSMeans. On the other hand, contrac- tors doing business with MnDOT showed a clear preference for Option 2 over the use of external existing indexes. While conducting the research on IDIQ methods for MnDOT (Rueda 2013), it was decided to take a closer look at the use of external indexes to adjust unit prices over time. To analyze this practice, 12 different cost indexes, including one published and maintained by MnDOT, were applied to four different types of projects over a five-year period. The calcu- lation of the difference between unit prices obtained by using these indexes and actual prices observed on MnDOT’s his- torical bid data showed a poor correlation between external cost indexes and actual changes in the construction market. Table 20 shows average variations obtained by each of these indexes on each type of project. It is important to note that these analyzes only apply to construction projects executed by Minnesota. These indexes may better represent construction price changes in other states. This poor correlation may also be perceived by contractors doing business in Minnesota, so that it could be the reason for preferring a fixed annual adjustment rate included in the bid package. What this analysis shows is that the applicability of external indexes in IDIQ contacts to adjust unit prices over time could to be assessed on a per

46 FIGURE 32 Price escalation approaches—Suitability diagram. agency basis. These indexes may not be equally applicable for all agencies. Given the previous analysis and the apparent acceptance of MnDOT’s contractors for an alternative approach as the one described in Option 2 (see Figure 32), a method called A times E (A × E) bidding was developed. Similar to A + B con- tracting, in an IDIQ A × E contract contractors are required to bid in two different parts; A and E. In part A, contractors must submit unit prices for those pay items and bid quantities advertised by the agency; items that are expected to be repeat- edly used in different work orders throughout the contract and bid quantities that are intended to be in proportion with typi- cal work orders. In part E, bidders are required to submit a fixed annual adjustment rate to be used to modify bid unit prices on the anniversary date of the letting of the contract. This adjustment rate is then transformed into an escala- tion multiplier (E), which along with the price proposal (A),

47 Cost Indexes Average Variation (+/-) Asphalt Pavement Concrete Pavement Traffic Barriers Drainage Average per Index Building Construction Indexes (National) RSMeans—CCI (National) 18.82% 7.93% 6.44% 10.83% 11.00% ENR—BCI (National) 18.76% 8.07% 10.25% 10.28% 11.84% Average per Type of Project 18.79% 8.00% 8.34% 10.56% — Highway Construction Indexes ENR—CCI (National) 17.20% 7.72% 11.07% 9.30% 11.32% BLS—PPI 26.98% 16.54% 10.62% 17.52% 17.91% NHCCI (National) 33.83% 25.16% 20.94% 26.41% 26.58% Caltrans (Quarterly) 30.12% 19.96% 26.47% 21.90% 24.61% Caltrans (12-M) 27.06% 17.59% 20.56% 18.94% 21.04% SDDOT 16.96% 6.48% 12.38% 8.15% 10.99% Average per Type of Project 25.36% 15.58% 17.01% 17.04% — Minnesota & Minneapolis Indexes RSMeans—CCI (Minneapolis) 18.33% 7.63% 11.02% 10.61% 11.90% ENR—BCI (Minneapolis) 19.96% 9.40% 9.96% 10.76% 12.52% ENR—CCI (Minneapolis) 20.34% 9.46% 10.26% 11.21% 12.82% MnDOT—CCI 18.09% 5.50% 12.92% 10.19% 11.68% Average per Type of Project 19.18% 8.00% 11.04% 10.69% — Source: MnDOT (2014). ENR = Engineering News-Record; NHCCI = National Highway Construction Cost Index. TABLE 20 AVERAGE VARIATION PER INDEX AND TYPE OF WORK compose the selection formula (A × E) used to determine the low bid. Because contractors are expected to make better predictions with regard to future construction prices (Rueda 2013), A × E bidding is expected to increase the accuracy of price escalation methods as well as increase the confidence of contractors in understandable, fair, and transparent escala- tion clauses. As a result of the lower uncertainty perceived by bidders, this method is expected to reduce bid unit prices by lowering the need for contingencies related to the use of inadequate price escalation methods and the higher risk inherent in long-term construction contracts. Tables 21 and 22 contain an example of the use of this method using a preliminary equation for the calculation of BID SCHEDULE Pay Items Quantity Bidder 1 Bidder 2 Unit Price Extension Unit Price Extension Item X 11,631 $0.63 $7,328 $0.90 $10,468 Item Y 1,479 $64.55 $95,469 $64.00 $94,656 Item Z 1,530 $2.72 $4,162 $2.82 $4,315 Total Bid $106,959 $109,439 TABLE 21 A × E EXAMPLE—BID SCHEDULE A Escalation Rate (r) E A  E Bidder 1 $106,959 8% 1.03%* $110,450 Bidder 2 $109,439 2% 1.01% $110,319** Eq. 1 A = Price Proposal; E = Escalation Multiplier * As shown in Eq. 1 ** Bidder 2 wins. TABLE 22 A × E EXAMPLE—CONTRACTOR SELECTION

48 the escalation multiplier (Equation 1 in Table 22). It was determined that this equation represents an acceptable level of risk for MnDOT in relation to effectively awarding the low bid. This equation corresponds to a weighted sum of bid schedules for a three-year period. A three-year period was used given that the maximum number of times MnDOT will adjust its IDIQ contracts during the base contract period (construction time without extensions) is two. About 60% of current MnDOT IDIQ contracts will be effective for at least three contract periods. No contract extensions were consid- ered when developing this equation because once finished with the base period, MnDOT will have the opportunity of deciding whether or not to extend the contract in accordance with adjusted unit prices at that moment, actual unit prices that would be obtained if reprocuring the contract, and the cost of executing a new contract. Weights assigned to the first, second, and third contract peri- ods are 70%, 20%, and 10%, respectively. Weights were deter- mined by considering the relative importance of each period. Lower unit prices are more important during the first period because the minimum guaranteed amount is usually covered during the first year of the contract. Once this minimum amount of work is reached, MnDOT could stop issuing work orders to the contractor on the second or third period. Likewise, obtain- ing lower prices for the second period is more important than getting those for the third period given the higher probability of performing work during earlier contract periods. It is important to note that this is a preliminary model and the escalation multiplier equation is to be modified as a result of an assessment of bidders’ reactions and behavior as a result of the implementation of this approach. In addition, this model was specifically developed for the MnDOT single award IDIQ contracts after a comprehensive analysis of its historical bid data; therefore, its applicability for other agencies should be carefully analyzed. Table 21 shows an example of a bid schedule with unit prices submitted by two different bidders for three pay items. In a traditional low-bid approach, Bidder 1 would be selected for the execution of the contract. However, Table 22 shows that the most effective overall proposal when considering all contract periods is the one from Bidder 2. INDEFINITE DELIVERY/INDEFINITE QUANTITY SURETY BONDS Developing pragmatic bond schemes for IDIQ contracts is an important issue given the typical high uncertainty in IDIQ techniques associated with actual quantities of work to be required under the contract. Usually agencies must select one of the following bond schemes: • Submission of performance bonds by contractors cov- ering the maximum amount to be ordered under the contract; • Contract bond at award covering the minimum guaran- teed amount or total bid schedule (list of bid unit prices and quantities) and subsequent bonds on a work order basis; or • Only bond each work order with no contract bond at award. The crux of the issue revolves around tying up an IDIQ contractor’s bonding capacity when there is no guarantee that the owner will be at risk for more than a single guaranteed minimum amount of construction placement. A recent study on performance bonding funded by FHWA (Kraft et al. 2014) found that small contractors and disadvantaged business enter- prises generally had little net working capital and, as a result, were often not able to arrange bonding for projects. In addi- tion, that study found that even when a small contractor could furnish the required bond, its bond premium could be as much as three times higher than an established contractor’s pre- mium, making it impossible to win a low-bid award (Kraft et al. 2014). The federal literature cited IDIQ contracting as an important tool for federal agencies and state agencies using federal-aid funding to meet their mandated small busi- ness goals because the work orders are typically less complex than a typical construction project and the repetitive nature of the delivery method allows the small business owner’s proj- ect management team to leverage the experience gained on one work order to their advantage on the next (Alinger 2010). This section presents and analyzes the opinions of differ- ent contract stakeholders obtained from the surveys, including one survey conducted with surety companies doing business in Minnesota. It is important to consider that most of these sureties also have business in other states. Additionally, 94% of 34 surety companies reported that they have furnished bonds for IDIQ contracts. Figures 33–35 show the survey responses obtained from state DOTs, AGC/ARTBA contractors, and MnDOT’s con- tractors. It can be noted that the question for contractors was stated differently than the way it was posed to DOT respondents. Unlike state DOTs, contractors were requested to select the least preferred alternatives. Because most sur- vey participants from both groups of contractors did not pro- vide their opinions for this question, it can be inferred that most contractors might not bid for IDIQ contracts if they were required to furnish a bond for the maximum contract amount. Conversely, most state DOTs appear to consider this scheme as the most suitable approach. During the study conducted for MnDOT and in order to obtain better insight into this issue, survey responses were received from 39 representatives of surety companies doing business in Minnesota. Figure 36 shows the opinions of these sureties with regard to the suitability of each bond scheme. Unlike state DOTs, most sureties prefer a bond scheme in which a contract bond is provided at the time of award cover- ing the minimum guaranteed amount with subsequent bonds

49 FIGURE 33 State DOTs—Suitable bond schemes. FIGURE 34 AGC/ARTBA contractors—Bond schemes. FIGURE 35 MnDOT contractors—Bond schemes.

50 (once covered the minimum guaranteed amount) on a work order basis. Several comments submitted by surety companies men- tioned the impact that using bonds covering the maximum amount of the contract would represent for small contractors; 82% of the sureties indicated that this approach would signif- icantly impact a contractor’s ability to bid for other contracts; having a higher impact on small contractors (Gransberg and Rueda 2014), validating the finding about small contractors in the FHWA performance bonding study cited earlier (Kraft et al. 2014). In addition to the amounts covered by the bonds, some surety companies identified longer contract periods as a fac- tor that may negatively impact a contractor’s ability to receive a bond. Some mentioned that IDIQ contracts for more than one or two years make bonding difficult for contractors. One oft-repeated comment found in this survey referred to the determination of maximum contract amounts by year, which is a practice used on FDOT DB-PB contracts. Among the comments submitted by surety companies, there was one that proposed a different approach that is worthy of consideration when selecting a bond scheme. It consists of setting a fixed bond amount to cover the own- er’s default exposure at any given time during the contract period. For example, if it is known that the contractor will not be performing, at any one time, work orders for more than $500,000, regardless of the maximum amount of the contract, the agency could set a bond for $500,000 or another amount large enough to cover this value (Gransberg and Rueda 2014). The FHWA performance bonding study found the default rate for DOT projects to average just 0.65%, and that average bond premiums for all U.S. contractors added 1.139% to the cost of every construction project (Kraft et al. 2014). In other words, given that the U.S. Census Bureau in 2014 estimated that public construction in 2013 was approximately $270 billion, the tax payer paid about $3 bil- lion that year to be protected from an event that happens less than 1% of the time. Therefore, an agency’s decision on how to approach performance bonding can potentially nega- tively impact small businesses in order to provide protection against a risk that is rarely realized. Thus, if a given IDIQ contract is intended to stimulate small business participation, minimizing the size of the required bond is in order to main- tain the deepest possible pool of competitors. The following statements summarize several oft-repeated comments submitted by surety companies with regard to IDIQ contracts. • Agencies are to limit the contract period to one or two years or allow surety companies to furnish annual bonds. • The contract includes maximum quantities of work (in dollars) to be assigned to contractors each year. • Agencies are to establish a maximum value of work orders (in dollars) that the contractor may be perform- ing at any one time. SUMMARY This chapter found that unlike multiple award contracts a single award approach permits a greater interaction between owners and contractors, allowing an early involvement of contractors in work order development procedures. This interaction is limited in multiple award contracts as a result of the implications of handling multiple contractors. How- ever, the use of the latter eliminates the use of mobilization and escalation clauses since price proposals are submitted by contractors on a work order basis. This chapter also highlights the risk associated with inap- propriately sized work orders. It was found that the issuance of work orders smaller than the usual size of a traditionally procured contract for the same type of work may encourage contractors to submit higher than normal bid pricing. Like- wise, work orders that are too large may prevent the agency from reasonably awarding future work. FIGURE 36 Surety companies—Bond schemes.