Cover Image

Not for Sale



View/Hide Left Panel
Click for next page ( 38


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 37
27 mance-based maintenance contracts. The role of PBMC in The cost savings for the 10-year, lump-sum, performance- publicprivate partnerships, such as concessions and multi- based contracts has been estimated at 10% (Stankevich et al. phase highway projects (for example, DBOM), has come to 2006). Pakkala (2002) asserted cost savings, but no quan- be fully appreciated (Pakkala 2002, 2007). titative estimate was provided. A regression analysis per- formed on a very small sample and including a statistically insignificant variable suggested that costs actually increased INTERNATIONAL EXPERIENCE (Stenbeck 2007). Canada Most of the former government employees have fared well working for private contractors (Stankevich et al. 2006). In Canada, the development of performance-based contracts However, the published literature reviewed does not provide for road maintenance started in the late 1980s and early data or quantitative analysis to support this assertion. 1990s. British Columbia piloted the first PBMC in 1988 (Zietlow 2005a). Performance standards still leaned toward Alberta required work procedures rather than outputs or outcomes. Today, British Columbia uses performance-based contracts In the province of Alberta, PBMC was an outgrowth of a pre- to maintain 100% of its provincial highways (Stankevich et mier's initiative to reinvent government. The Alberta Infra- al. 2006). structure and Transportation Agency divided the province into 30 MCAs. Then the agency began outsourcing mainte- British Columbia nance based on unit price (Pakkala 2002; Lali 2007). Origi- nally a contractor was limited to working in three MCAs but British Columbia embarked on PBMC as part of a program is now allowed to work in seven. Substantial effort is devoted of extensive and rapid privatization governed by a number to ensuring that both competition and contractor capacity of principles and policies. The private sector would deliver will be adequate. Contractor performance is judged against services if it is more efficient and the public interest is not target LOS. Service levels include both technical criteria and adversely affected, and no contractor can hold an interest in response times. Failure to meet the contract performance more than three maintenance areas. Government employees criteria results in penalties. were given a number of options: resign, transfer to a posi- tion in another part of government, accept early retirement, In Alberta, the contract term is 5 years with a renewal or accept employment with firms awarded the maintenance term of 1 to 3 years. Price receives the majority of the weight contracts. About 2,300 employees chose employment with in the contractor selection criteria (Stankevich et al. 2006). successful contractors, and relatively few picked the other options. The government sold off its equipment, so there was In Alberta, the original objective was to provide a LOS no turning back (Ribreau 2004). that was at least equal to the LOS in-house maintenance staff achieved but at a lower cost (Bucyk and Lali 2005). Since 2003, British Columbia has made lump-sum awards for performance-based maintenance contracts with 10-year The Alberta Infrastructure and Transportation Agency performance periods. The contracts address maintenance will not disclose information on the LOS contractors achieve. and repairs and do not include resurfacing, rehabilitation, This information is used to evaluate contractor performance and reconstruction. The services include surface mainte- and is an important factor in making contract awards. nance, winter maintenance, drainage, landscaping, struc- tures maintenance, sign work, emergency maintenance and Table 8 shows various estimates of cost savings and cost repairs, and fixing damage to government property. A vari- increases in Alberta. ety of work is quantified and serves as the basis for perfor- mance measurement and incentives. Performance standards Ontario are proactive and customer-oriented. The transportation agency does not prescribe the methods for doing work. Each The Ontario Ministry of Transportation (MTO) has defined month contractors receive one-twelfth of the annual lump- Area Management Contracts that cover 60% of the provincial sum award provided that all performance standards are met; road network. Performance-based maintenance contracts are otherwise, deductions occur. 95% lump-sum and contract terms range from 7 to 9 years (Skinner 2007). The contracts include all routine mainte- A qualitative conclusion is that the LOS are at least as nance such as pothole repair, vegetation management, bridge good as they were before embarking on the performance- maintenance and cleaning, electrical work, and line painting. based contracts (Pakkala 2002). The public appears satisfied Other types of work addressed include winter maintenance, with the outcomes and the role the private sector is playing patrolling to conduct visual inspections, and emergency in maintaining the roads (Stankevich et al. 2006). assistance to deal with accidents and spills. Maintenance

OCR for page 37
28 performance standards include both outcome and time-based was responsible for overseeing the pilot projects. Oversight performance criteria. Failure to meet the standards can result responsibilities included inspection, prioritization of needs, in penalties. Over time, the duration of the contract period definition of work, design of treatments, work management, has been increasing gradually and the number of maintenance and reporting, among other things. The exact same docu- activities has grown (Stankevich et al. 2006). mentation and reporting procedures were used in each pilot. Next the RTA engaged a maintenance contractor and the in- One source indicates that Ontario has experienced better house forces to maintain their respective 100 km sections LOS; the network quality is at least as good as it was before of road (Stankevich et al. 2006). The RTA determined that, the performance-based contracts. No quantitative informa- during the first year, the private entity was able to achieve, tion was offered to support this assertion (Pakkala 2002). relative to the state forces, a 16% savings in costs, a 22% A knowledgeable staff member of the MTO said that infor- improvement in productivity, and a 13% improvement in the mation on the effect of contractors on LOS is not generally condition of assets (Segal et al. 2003). published (S. Skinner, personal communication, May 2007). As to cost savings, one of the sources stated that the MTO These findings encouraged New South Wales to seek bid- experienced some cost savings but did not provide backup ders and award a 10-year, $130 million performance-based information (Pakkala 2002). contract covering all maintenance activities for 450 km of Table 8 COST SAVINGS OF PERFORMANCE-BASED MAINTENANCE CONTRACTS IN ALBERTA, CANADA Contract Description Source Cost Savings Two rounds of tenders of hybrid contracts eventually covering 100% 5% savings for first round of tenders; (World Bank 2006) of provincial roads in 30 Contract 25%35% savings for second round of tenders Maintenance Areas (CMAs) Clients receive some cost savings (did not provide any quantitative (Pakkala 2002) estimates or supporting data) An independent KPMG report stated there were cost savings based on tendering into 17 CMAs: a 28% reduction in cost between the new (Bucyk and Lali 2006) contracts (year 2000) and the old contracts (prior to 2000), which translates into a reduction in unit prices to $3,705/km from $5,117/ km, representing a total annual cost reduction of $26,419,932. Included Alberta in a regression analysis of change in costs per kilo- meter owing to PBMC for southern Canadian provinces, Washing- Whole province (Stenbeck 2007) ton State, and Sweden. Determined that PBMC resulted in an increase in costs for all these jurisdictions. (See text for discussion of validity of this result.) Note: The base against which the cost savings are estimated is often not stated in the source. Australia urban roads (or 1,900 lane-km) in Sydney. The most important performance criteria were average roughness and cracking. Australia is composed of six states and two territories, The World Bank reported a 13% improvement in condition a number of which have been among the world leaders in accompanied by a cost reduction in the 20% to 30% range, using performance-based maintenance contracts. One of but it is not known over how many years of the contract these the first jurisdictions to try this type of contract was New results apply. The bid price was 25% lower than estimated South Wales. In 1990, this state began a comparative study (Stankevich et al. 2006). The Reason Public Policy Institute of two 100 km pilot projects in Sydney. According to the reported that since New South Wales started this performance World Bank, the objective of the study was to determine the contract, roadway condition has improved approximately 15% feasibility of contracting road maintenance and to estimate and there has been a 35% cost savings (Segal et al. 2003). differences in cost, quality, and responsiveness between a contractor and the RTA's workforce (Stankevich et al. 2006). Since then several new contracts have been let in New State forces performed the maintenance work in one pilot South Wales, Tasmania, and Southern and Western Austra- project and a private contractor conducted the maintenance lia. A number of them are hybrid contracts, in which some of work in the other. Results also were compared with work the maintenance is paid based on both quantities and perfor- done by the in-house force. The RTA of New South Wales mance standards (Zietlow 2005a). Table 9 shows estimated first awarded a management contract to a contractor that cost savings for various jurisdictions throughout Australia.

OCR for page 37
29 Table 9 COST SAVINGS FROM PBMCs IN AUSTRALIA Jurisdiction(s) Source Reduction in Costs Regional Transport Authority, New South (Pakkala 2002) 10% to 35% (bases for comparison not known) Wales; Tasmania; Western Australia Sydney, New South Wales (Frost 2001) 38% cost savings compared with schedule of rates type of contracts Southern Tasmania (Frost 2001) 20% cost savings compared with schedule of rates type of contracts South Perth (Frost 2001) 25% cost savings compared with schedule of rates type of contracts Mid North Region (Frost 2001) 30% cost savings compared with schedule of rates type of contracts Savings of 15% to 20% against other forms of maintenance contracting, which in turn were reported to have achieved 20% savings against in- Six contracts in Western Australia (Frost 2001) house operations. Thus, total savings compared with in-house operations are estimated to be at least 35%. Note: The base against which the cost savings are estimated is often not stated in the source. Table 10 COST SAVINGS FROM PSMCs IN NEW ZEALAND Contract(s) Source Savings 10-year, lump-sum, performance-specified mainte- nance contracts on part of the national road net- Reason Public 20% savings based on regular audits work and highway works throughout country; Policy Institute includes rehabilitation and maintenance World Bank Less cost according to General Manager of Transit New Zealand 30% decrease in cost of professional services and 17% Highway Mainte- decrease in costs of professional services; a savings of at least nance Contracting 25% over conventional model 10-year, lump-sum, performance-specified mainte- Initial savings were about 25%, and were between 14% and Pekka Pakkala nance contract (PSMC-001) covering 450 km 20% at time report was written. Savings predicted to be 25%. Note: The base against which the cost savings are estimated is often not stated in the source. New Zealand both method and performance specifications. A large num- ber of the performance standards are expressed in terms of In New Zealand, the national road agency is known as Transit intervention times. The term of the hybrid contracts is only New Zealand (Transit NZ). In 1998, it let its first long-term 5 years (Stankevich et al. 2006). performance-based maintenance contract known as a Per- formance-Specified Maintenance Contract (PSMC). Today, Industry experts have asserted that these PSMCs have lump-sum PSMCs with 10-year terms are used on 15% of the resulted in improved maintenance service and road quality. nation's entire road network, mainly on national roads. Con- The general manager of Transit NZ reported that better services tractors must satisfy a detailed set of key performance indi- were delivered. It is unclear whether this statement applies to a cators. Recent PSMCs measure performance at three levels: hybrid contract (with both performance and method specifica- management, long-term, and operational measures. The first tions) or to a long-term performance-based maintenance con- set of performance measures concerns contract manage- tract. An expert on PBMC wrote that there were pronounced ment and implementation. The second set--the long-term improvements in the quality of service, bumps, skid resistance, measures--pertains to the condition of the pavement and signs, drainage systems, and market posts for a well-known addresses such attributes as roughness, texture skid, and performance-based contract (Pakkala 2002). structural integrity. The third set--operational measures-- includes the condition of the appurtenances along the road, Various sources have consistently reported cost savings the effect on serviceability, and the user's experience. Tran- for New Zealand's performance-based contracts. Example sit NZ also has hybrid contracts that incorporate features of statements appear in Table 10.

OCR for page 37
30 United Kingdom it is difficult at this time to determine the effectiveness of the MACs. The author wrote that there is some overall evidence The U.K. Highways Agency is responsible for 8,850 km of that service has improved with slightly reduced costs, particu- the most strategically important part of the nation's highway larly with respect to routine maintenance (Harding 2005). network. This portion of the highway system amounts to only 4% of the total roadway miles in the United Kingdom, but it Finland carries 30% of the total traffic and 60% of the truck traffic. In 2001, Finnra undertook a major reorganization. The part The United Kingdom implements four models for road responsible for design, engineering, and maintenance and maintenance, ranging from pure method-based to pure out- operations was transformed into a wholly state-owned pro- come-based with different levels of flexibility. Maintenance duction organization known as the Finnish Road Enterprise currently occurs under three different models, but a fourth, (FRE) to compete with the private sector regarding capital, which has a financing component, is being developed. In maintenance, and operations projects. Finnra remained the one model, the Highways Agency enters into a contract client organization responsible for procuring contractors and with a consultant, the Managing Agent whose responsibili- entering into contracts. Initially, public tendering began in 23 ties include providing advice on procurement, procuring the of 99 maintenance areas on the network. Because of the poten- Term Maintenance Contractor (TMC), providing a long-term tial impact of the FRE on the competitive position of existing focus, performing engineering, and assisting in strategic private firms, the FRE was introduced gradually into the mix matters including promoting innovation. Once the Highways of contractors. Full and open contracting involving the FRE Agency and TMC have entered into a contract, the Manag- throughout the country did not start until 2005. The origi- ing Agent provides instructions to the TMC regarding what nal 23 maintenance area contracts were lump-sum, mainly maintenance work to perform. Under a certain threshold, the output (accomplishment). A few were outcome-oriented, and TMC may perform rehabilitation work and upgrades. these were 3-year contracts. The contractor selection criteria were based 75% on price and 25% on technical qualifications Under a second model, a Network Board is established (Pakkala 2002). to provide further strategic direction, strengthen partnering, monitor highway improvements, and perhaps resolve con- As of approximately 2005, Finnra maintenance contracts flicts or complaints. were hybrids, 75% of compensation based on a lump-sum payment with adjustments for failing to meet performance The third model is administratively the simplest and criteria, and 25% based on unit price. The outcome-based transfers management of the network to a Managing Agent portion of the contract, covered by the lump-sum payment, contractor while creating a Network Board to provide stra- concerned winter and summer maintenance, minor bridge tegic direction. maintenance, gravel roads, vegetation management, and drainage and culvert maintenance, among other things. Performance-based maintenance contracts are known as For example, intervention time and outcome specifications Managing Agent Contracts (MAC) in the United Kingdom. for winter maintenance required the contractor to respond These contracts incorporate performance specifications to to snow conditions within 2 hours, remove all snow to no increase efficiency and effectiveness, allocate responsibility more than 1 cm within a certain amount of time after it stops and risk between the client and the contractor, foster innova- snowing, and achieve a measure of skid resistance of less tion, and focus the attention of the contractor on outcomes. than 0.3. Some important characteristics of MACs are as follows: Of the remaining 25% of compensation for area con- Increased outcome orientation tracts, payment is based on unit prices. Some of the types Lump-sum and unit prices as a basis for payment of work compensated in this fashion include the renewal of Strengthened partnering roadsides, replacing guardrails, and drainage repair. Emphasis on continuous improvement Focus on life-cycle costs Finnra enters into separate unit price contracts to address Better risk management lighting, road markings, traffic signs and signals, resurfac- Supply chain management (Pakkala 2002; Harding ing, and rehabilitation (Stankevich et al. 2006). 2005). The major documents about Finland's experience with Little is published or easily accessible regarding changes PBMC do not contain quantitative information and trends in LOS or costs regarding PBMC in the United Kingdom. An regarding changes in LOS (Pakkala 2002, 2007). Table article published in the proceedings of a seminar on the state 11 summarizes estimated cost savings from some key of the practice regarding PBMC around the world reported that sources.

OCR for page 37
31 Table 11 features, and aesthetics including vegetation. Penalties were COST SAVINGS FROM PBMCs IN FINLAND imposed gradually to give the contractor an opportunity to address the deficiency, but would reach 100% of the penalty Source Savings if the problem was not redressed by the third inspection at Finnish Road Enterprise Cost savings analyzed at 7% to the end of the month. (Pakkala 2005) 10% for 3-year contracts and 13% for 7-year contracts These kilometer/month performance-based maintenance World Bank (Stankevich 7% to 10% for 3-year contracts and contracts were judged to have worked well. The average cost et al. 2006) 13% for 7-year contracts; the cur- of routine maintenance was $175/km/month and penalties rent price level is 50% to 60% of resulting from 600 instances of noncompliance totaled only the price level when Finnra was $300,000, approximately 1% of the value for all the contracts using its own labor and equipment to do maintenance (Cabana et al. 1999). World Bank Transport 30% to 35%; about 50% less cost/ The kilometer/month contracts were followed by a new Note TN-27 (Stankevech km and innovative contract designed for combined rehabilita- et al. 2005) tion and maintenance of paved roads. This new contract- Note: The base against which the cost savings are estimated is usually not stated in the source. ing process, CREMA, unfolded in two phases. In the first phase, CREMA required the contractor to rehabilitate and then maintain a network of 11,818 km of nonconcessioned Latin America roads for five years for a lump-sum amount. Each contract covers contiguous segments with lengths ranging from 100 The average condition of one-third of the roads in developing km to 300 km and specifies the road sections that require countries is poor. In Latin America, the condition of roads rehabilitation. Contracts tend to exceed 10 years. A mini- is even worse. A few years ago it was reported that only 7% mum overlay thickness was specified to ensure a positive to 52% of roads were in good shape. Even in the best-case net present value given a 12% discount rate. About half the scenario, nearly half of the roads are in poor shape (Segal et contract funds were paid to the contractor during the first al. 2003). In an effort to improve road conditions and reduce year to cover the rehabilitation costs and the balance was maintenance costs, Argentina and Uruguay were pioneers in paid as a lump-sum spread equally over 48 months. Deduc- adopting performance-based contracting models. tions for failing to satisfy both technical and timeliness standards were imposed. The technical performance stan- Argentina dards pertained to visual deterioration, potholes, cracks, rut- ting, blockage of drainage systems, friction, and deflection Since 1995 Argentina has pursued a number of different (Cabana et al. 1999). contracting approaches involving performance-based main- tenance, including kilometer/month contracts for routine The second phase of CREMA addressed 8,200 km of maintenance, Contrato de REcuperacion y MAntenimiento nonconcessioned roads under combined rehabilitation and (CREMA) Phase I, CREMA Phase II, and concessions. maintenance contracts totaling $550 million. Upon comple- tion of the original rehabilitation work, contractors were Argentina's national system of roads totals 38,744 km, required to maintain the roads in accordance with technical of which 30,912 km are paved and 9,508 are concession toll and timeliness specifications as in CREMA Phase I. These roads. In 1995, Argentina launched a series of performance- roads were in worse condition than those addressed in the based maintenance contracts covering about 3,600 km of first phase. Thus, it was necessary to rehabilitate a higher these national roads. The government selected paved roads percentage of roads--in the range of 65% to 70% in Phase in fair to good condition for 11 maintenance contracts, each II compared with 40% in Phase I. In Phase II the contractor varying from 105 km to 536 km for a total of US $650 mil- was required to perform rehabilitation using thicker over- lion. Contracts were lump-sum and payments were made lays, more in line with the optimal thickness determined by each month according to an amount per kilometer (based the World Bank HDM model. Also in Phase II, the period in on equivalent liters of gasoline to account for inflation but which rehabilitation could occur was increased from 12 to expressed here as dollars/kilometer/month). The contracts 24 months or longer (Stankevich et al. 2006). were for 2 years and were renewable. Three inspections, based on a sampling process, occurred once per month. If Under another contracting approach, to obtain additional one or more deficiencies was found with regard to perfor- funds to finance the rehabilitation and maintenance work of mance targets--both technical specifications and response medium- and low-volume roads, the government awarded times--the road agency made deductions from payments in 12-year concessions that permitted contractors to collect accordance with a table of penalties. Performance standards tolls. The government also made payments to the contractor addressed specific measures concerning ride quality, safety for additional costs (Segal et al. 2003).

OCR for page 37
32 Because the first phase of CREMA performance-based Uruguay contracts began with rehabilitation followed by maintenance, it had a significant effect on asset condition. The percent of Beginning in 1996, the Ministry of Public Works introduced roads in good to fair condition increased from 59% to 94%. performance-based contracts for the maintenance of Uru- The percent in critical to poor condition declined from 41% to guay's national highway network. There were essentially 6%. Also, roughness measurements deteriorated at a slower two types of contracts; one concerned just routine mainte- rate than predicted by the HDM (Stankevich et al. 2006). nance, and the other involved initial rehabilitation followed by periodic and routine maintenance. A report of the Reason Public Policy Institute states that Argentina has achieved reduced costs of maintenance, better The first type of contract (for routine maintenance) gave LOS, new capital investment in the highway network, and a the employees of the ministry an opportunity to start their reduction in the government's maintenance workforces; it own contracting businesses. At the same time, the ministry also largely eliminated corruption in highway maintenance was able to reduce the number of its employees. To reduce and rehabilitation (Segal et al. 2003). However, an interna- the anxiety of staff and provide an inducement to work for tional expert on PBMC wrote that until now only the con- private contractors, staff could return to the ministry if the tracts in Australia, New Zealand, and the United States have contractor failed during the first year. The contracts were all reported substantial cost savings, whereas in Latin America successful. More staff desired to participate in the new con- no comparable cost analysis has occurred (Zietlow 2005). tracts than the contracts were able to absorb. A more recent resource guide of the World Bank concluded that the first phase of CREMA was determined to cost 16% Because the new contracting approach worked so well, the to 20% more than conventional contracting, but adjustments ministry carefully planned and piloted a second type of con- were not made to put the highway agency and contractors tract: a performance-based contract similar to Argentina's on a level playing field. An analysis after the fact suggested that involved both rehabilitation and periodic and routine the rate of return on rehabilitation and maintenance work maintenance. Rehabilitation occurred on selected sections was 60% calculated at a 12% discount rate. The lump-sum of highway before maintenance occurred. By the start of contract arrangement during Phase I virtually ensured no 2000, more than 40% of the national road network was being overruns except for force majeure events. Annual costs maintained under 5-year, performance-based maintenance never exceeded the lump-sum amount by more than 3%. contracts (Zietlow 2005b). Finally, the 60 Phase I CREMA contracts were judged to be financially attractive to the private sector, because only one The first performance-based maintenance pilot in Uruguay contract was cancelled (Stankevich et al. 2006). resulted in an improvement by two contractors in road condi- tions from 1996 to 1998. The first contractor increased roads Other conclusions of the World Bank resource guide in very good condition from 0% to 25% and reduced roads in regular condition from 40% to 15%. Roads in good and regarding Phase I of the CREMA include the following: bad condition remained unchanged at 60% and 0%, respec- tively. The second contractor increased roads in very good This lending institution was able to disburse its funds condition from 23% to 37% and those in good condition from much more quickly to rehabilitate and maintain the 13% to 46%, and reduced roads in regular condition from highway sector. 64% to 17%. The percent of roads in bad condition remained unchanged at 0% (Stankevich et al. 2006). Lump-sum contracts essentially eliminated cost overruns. The government was forced to bear long-term obliga- City of Montevideo tions to pay the contractors. Supervision costs fell sharply because the contractors In 1996, as Uruguay started transforming its road contract- supervised themselves. ing process, Montevideo commenced its first performance- The relatively simple set of performance indicators based contract for nearly 140 km of its city roads. The city permitted a small inspection team to monitor the entered into a 3-year contract with a 3-year extension. Por- contractors. tions of the network needed initial rehabilitation. Montevideo The performance specifications, which focused on out- paid the contractor based on unit price for the rehabilitation comes, fostered innovation. work. During the 3-year extension, the contract called for The likelihood of poor quality rehabilitation work was cutting monthly payments by 40% because of the rehabilita- minimized because of the contractor's responsibility to tion work that previously occurred. maintain the roads over the next 4 years and the penal- ties for failing to satisfy the performance standards. The city defined performance standards, response times, Capital improvement needs were estimated to decline and penalties for noncompliance for pavements, shoulders, by 30% (Stankevich et al. 2006). and drainage systems. Because road conditions at the start