Mr. Melia works for the Pioneer Institute for Public Policy Research in Boston, Massachusetts.
Public transit agencies across the nation are struggling to control costs without reducing service. One effective strategy used by a growing number of agencies is to contract privately for bus operations. Almost without exception, these agencies report that private bus companies can deliver equal or better service at a 20 to 30 percent lower cost.
The Massachusetts Bay Transportation Authority (MBTA) wants to use the same strategy and has accepted two bids to run 40 percent of its bus operations. The MBTA's cost analysis shows that it would save $23.1 million over five years. However, a 1993 anti-privatization law (commonly called the Pacheco Law) sets up a series of tests that an agency must pass before it can award a contract to a private company. Among the Law's more significant provisions are the following:
· Contractors must pay wages that are not lower than the lesser of these two:
(a) the minimum wage rate paid to state employees, or
(b) the average private sector wage rate for comparable positions, provided that such a rate has been established by the Executive Office for Administration and Finance;
· The private bid must be compared not to the existing cost of providing the service but to the cost that would result if public employees were working in the most cost-effective manner possible; and
· The contracting agency must demonstrate that the quality of service likely to be provided by the private contractor will equal or exceed the quality that "could be" provided by public employees, not the existing quality of service.
The Law gives the State Auditor the authority to decide whether these and other provisions of the Pacheco Law have been met, and to issue guidelines directing how state agencies must apply the Law.
The Auditor rejected the MBTA's proposed contracts, claiming that they "will not result in a cost savings" and that the MBTA "does not know whether the proposed contractors can meet acceptable quality service levels." The MBTA has sued, claiming that the Auditor applied the Law in an "arbitrary and capricious" manner, and that the Law itself is unconstitutional.
In preventing the MBTA from privately contracting for bus service, the Auditor did not develop his own comparison of the MBTA's in-house cost versus the cost of the private bids; nor did he provide any evidence that the service to be provided by the contractors is likely to be inferior to the current level of service. While the Auditor has identified several mistakes in the MBTA's cost analysis, these mistakes are relatively small and do not justify denying the contracts. Even if his objections were correct, the larger of the two contracts (which this paper analyzes in detail) would still save a minimum of $5.2 million over current costs.
This dispute between the Auditor and the MBTA reveals several important weaknesses in the Pacheco Law. On the surface, the Law seems to call for an objective cost analysis, but in effect, the Law allows the Auditor to deny a contract based on highly subjective judgments.
As it stands, the Pacheco Law prevents all but the most determined agency managers from even attempting to contract privately for services. (In the three years before the Law was enacted, there were 36 contracts awarded to private companies to provide a service formerly provided by state employees. Since the Law was enacted, five contracts have been awarded, only one of them significant.) Because competitive contracting usually results in a significant increase in productivity and cost savings, the public interest would best be served by amending the Pacheco Law to eliminate its most undesirable and arbitrary features. I recommend the following:
· Narrow the scope of the review to cost. It is difficult to predict quality in advance. Allowing the Auditor to speculate on the quality level of the services to be provided by the contractor is of no value to the public, especially when the Auditor lacks expertise in the service being contracted.
· Compare the private contract proposal to existing in-house costs, not the cost that "could be" achieved if the state agency worked in the most efficient manner possible.
· Require that the same assumptions the state uses when forecasting tax collections be used when preparing estimates of future variables such as wage increases and interest rates. This would remove the opportunity for either the Auditor or the contracting agency to determine the outcome of the review by manipulating important underlying assumptions.
· Require agencies to publish cost allocation plans as part of their annual budget requests. For agencies desiring to contract out, this requirement would remove much of the temptation to choose a cost methodology that would "prove" that a private contract would save money.
· Limit the Auditor's back-end review to identifying clear mistakes made by the contracting agency. The Auditor should not be allowed to raise generic doubts about an agency's accounting system or cost allocation methodology. (These issues should be worked out by the contracting agency and the Auditor before an agency solicits bids. Doing so should prevent either side from choosing a methodology simply because it would help "prove" their view of whether the contract will or will not save money.)
· Require the Auditor to come to a firm conclusion. If he finds mistakes, he must quantify them and factor them into the analysis. Then he must state what he believes is the in-house cost, and what he thinks would be the cost of the contract.
Bus transportation is a vital public service. The MBTA recently announced a cost-neutral plan to maximize bus ridership: they will reduce service on certain bus routes in order to increase service on more heavily used routes. In contrast, many other transit agencies have discovered that privately contracting for bus service allows them to run more buses and still cut costs. Unfortunately, the Pacheco Law will prevent Massachusetts from testing that option.
Public transit ridership in Boston has increased steadily since 1975, and the transit system has been both extended and modernized. In recent years, maintaining or increasing the level of transit has become more important, both to mitigate the impact of the Central Artery construction project and to help spur the development of the South Boston waterfront. The Massachusetts Bay Transportation Authority (MBTA) has recently extended commuter rail service to the South Shore, and plans to extend service to Worcester.
Yet there is a problem—the MBTA has long been one of the costliest transportation agencies in the nation. The Federal Transit Agency (FTA) reports it costs the MBTA $95 to run a bus for one hour. The median cost for the 25 largest U.S. transit agencies is $72 per hour, making the MBTA 32 percent more expensive than average.1
The MBTA has opted to maximize ridership by keeping fares low and extending service. At 85 cents for the subway and 60 cents for a local bus ride, the MBTA's fares are among the lowest in the nation, and are 30 percent lower (in real terms) than they were in 1965. Keeping fares low is a time-tested way to maintain ridership.
As affluence increases, so does suburbanization and the rate of automobile ownership. One study concluded that ridership would have decreased at least 10 percent between 1970 and 1990 if MBTA had not reduced the cost of fares and improved service.2 Today, revenues cover just over one-third of the agency's annual operating costs. When capital costs are included, the fare recovery ratio sinks to just 22 percent.
In 1965, the MBTA's first full year of operation, its deficit (the difference between revenues and the annual operating plus debt service costs) was $20.8 million. By 1991, the deficit totaled $575 million. In fiscal year 1998, the deficit will approach $675 million.3 One transit analyst estimates that the deficit will grow to $1.62 billion by the year 2010, assuming the MBTA continues to follow the same fare and service policies.4 That would require a subsidy of more than $4,500 per commuter per year.5
The deficit cannot forever grow faster than inflation (or tax revenue), and there are four basic strategies to reduce its growth rate:
· cut service,
· raise fares,
· drastically reduce automobile usage, and
· cut the unit costs of running trains and buses.
Cutting service is extremely difficult to do. When faced with periodic budget squeezes, MBTA management inevitably floats the idea of cutting the least-used bus routes. But it is easier to close a military base than to cut a route where the bus is always nearly empty, because each bus route has a very vocal constituency.
Raising fares is also unlikely to help contain the deficit. The MBTA is required to conduct an environmental impact study before it can implement a fare increase, and, if the study shows that there will be a significant loss of ridership, the MBTA must abandon or reduce the fare increase.6 Because ridership decreases about 1 percent for every 4 or 5 percent increase in fares, any fare increase large enough to have substantial impact on the deficit would almost certainly trigger a significant reduction in ridership .
Several European and Asian cities are using advanced technology (e.g., smart cards, electronic tolls) to charge drivers a premium for entering the central city during rush hour. These experiments have shown that if the marginal cost of driving during rush hour is high enough, automobile usage can drop 20 percent or more. With such policies in place, transit agencies can raise fares without fear of losing riders. However, given the public response to recent efforts to reduce or eliminate tolls on the Massachusetts Turnpike, such a policy is unlikely to be adopted in Massachusetts anytime soon.
There is, however, some good news. Evidence suggests that controlling unit costs can help curb the deficit. If the MBTA can become more efficient, it might be able to reduce its costs without cutting service or raising fares. Over the past 15 years, a number of European and American cities have achieved significant cost savings by contracting privately for bus service. The MBTA is now attempting to do the same. In December 1996, it received bids to run about 40 percent of the bus service, and calculates that the two selected bids would save the agency about $23.1 million over the five-year term of the proposed contracts.7
Nevertheless, under Massachusetts General Laws (Chapter 7, Sections 52-55, the so-called Pacheco Law, named after its primary sponsor, Senator Marc Pacheco), the MBTA is subject to the following requirements:
· the MBTA must determine its in-house cost of performing the bus service;
· the MBTA may award the contracts only if a private company can provide equal service for less money, or better service for the same or less money; and
· the State Auditor must then review the MBTA's analysis and either verify that the terms of the Law have been met, or point out where the MBTA has not complied with the Law.
This appears to be a fairly straightforward process. To determine the cost of providing the bus service in question, the MBTA would do the following:
· identify all of its direct costs (e.g., wages, fringe benefits and uniforms for bus drivers and mechanics; diesel fuel; tires);
· allocate a reasonable portion of indirect costs (e.g., management costs; support costs such as payroll and human resources); and then
· compare this total to the outside bids.
The process became ensnared during the first year after the bids were received when the following events occurred: the Auditor accused the MBTA of deliberately withholding and distorting information, thereby making it impossible for the Auditor to conduct a comprehensive and accurate review; and the MBTA General Manager accused the Auditor of willfully ignoring his explanations. The Auditor and the MBTA have disagreed over virtually every aspect of the cost analysis. The MBTA claims the two contracts8 would save at least $16 million, even if every one of the Auditor's criticisms were correct. The Auditor determined that "contracting out this service will not result in a cost savings."9
Furthermore, the Auditor has ruled that the proposed contracts are not in the public interest, and has prevented the MBTA from awarding the contracts. In response, the MBTA sued, claiming the Auditor's review was "arbitrary and capricious."10
This dispute is significant not only because millions of taxpayer dollars and hundreds of union jobs are at stake, but because it is seen as a test case. If the court upholds the Auditor's decision, new competitive contracting by state government is essentially dead in Massachusetts. If the court overturns the Auditor's decision—especially if the court finds the Law to be unconstitutional—then competitive contracting may expand rapidly, likely becoming central in MBTA's efforts to maximize ridership without imposing an intolerable deficit on the taxpayers.
The purpose of this paper is to determine whether the Pacheco Law created an objective framework for assessing the value of private contracting, which framework would then allow the state government to make decisions regarding the most cost-effective means of providing services to the public. The author examines the role of the Auditor in this process, and analyzes the MBTA's responses to the Auditor's objections. The paper concludes with recommendations for improving the process by which private contracting proposals are evaluated.
To understand how this conflict arose, three issues must be examined:
· the MBTA's history of cost control efforts;
· the experience of other transit agencies with privately contracted bus service; and
· the politics surrounding private contracting in Massachusetts.
A. The MBTA's Cost Control Efforts
The MBTA's costs spiraled out of control during the 1970s. Jose A. Gomez-Ibanez, a transportation analyst at Harvard University, found that "real unit operating costs increased rapidly in the 1970s, accounting for 31.1 percent of the deficit growth in that decade."11 Because the MBTA's costs grew faster than inflation, Gomez-Ibanez determined that in 1980, the MBTA's operating costs were $29.2 million higher than they would have been had there been no growth in real unit operating costs. With a total operating budget of less than $275 million, this decline in productivity consumed more than 10 percent of the MBTA's 1980 operating budget.
By 1981, the MBTA plunged into a fiscal crisis. The 78 cities and towns served by the MBTA demanded relief. Fares were doubled (but then partially rolled back), and the annual increase in the contribution that any city or town had to make toward the MBTA's deficit was capped at 2.5 percent, with the state paying the excess. To control costs, the legislature enacted the Management Rights Act of 1981. This law gave MBTA managers significantly stronger cost control tools, including ending automatic cost-of-living adjustments, eliminating pensions on overtime, allowing the use part-time employees, and giving the option to contract privately for services that had traditionally been performed by MBTA employees. Six years after the legislation, the MBTA estimated cumulative savings at $118 million.12
The Gomez-Ibanez analysis confirms that determined efforts to control costs can make a difference. His study found that real unit operating costs declined in the 1980s, in large part because of the use of the Management Rights Act. While the total deficit grew during the 1980s because of inflation, fare erosion, and a major capital improvements program, Gomez-Ibanez estimates that by 1990, the MBTA's cost control efforts were saving $49.5 million per year.13
B. Privately Contracted Bus Service
The private bus industry in the United States operates more than 120,000 vehicles14 and is beginning to run a significant number of buses under contract to public transit agencies. In the mid—1980s, only 1 to 2 percent of public transit bus routes were privately contracted, but by 1997, 10 percent of all routes were run by private contractors.15
In the United States, San Diego pioneered private contracting in 1979. Denver and Los Angeles have been using private contractors for about a decade. More recently, Indianapolis moved to contract privately for most of its bus service, and Las Vegas became the first transit agency to contract privately for all of its bus service. Private contracting is also becoming common in Europe, with London being the first major city to rely on it extensively.
In virtually every case, the use of private contracting allowed transit agencies to reduce their unit costs and increase the level of bus service, as Table 1 shows.
Cost Savings for Cities Contracting Privately for Bus Service
|Metropolitan Area||Percent of Bus
Service under Private Contract
|Percent Increase in Service|
Public transit managers have been very careful to set quantitative performance measures for private contractors. The two most common measures are the percentage of scheduled trips run and the average miles between breakdowns. Most public transit agencies also require private contractors to meet additional quality standards related to cleanliness, safety and passenger satisfaction. Follow-up studies in London, Denver and Los Angeles have all concluded that the quality of service is equal or better after private contracting.1
After 18 years of experience with private contracting, the verdict is clear: public agencies that contract privately for bus service can expect to save 20 to 30 percent and maintain a comparable quality of service. Most cities that have tried private contracting are pleased with the results and intend to expand its use. The Los Angeles official in charge of contracting for bus service has remarked,
For larger transit agencies, there is no one program or project that has the potential to decrease costs and improve service quality more than competitive contracting ... From a cost savings perspective, I have seen enough bids and been involved in enough projects to know that [one] could open a private sector bus operating company tomorrow and save large transit agencies at least 25 percent.17
C. The Politics of Private Contracting in Massachusetts
The very thing that makes private contracting attractive to transit agencies—cost savings of 20 percent or more—also makes private contracting controversial, because those savings ultimately come at the expense of labor. The two primary ways that private companies reduce costs are by improving productivity and by reducing wages and benefits. Improving productivity results in fewer transit jobs being necessary to provide the same level of service. Over the years, public transit agencies have entered into collective bargaining agreements containing restrictive work rules. These rules frequently limit the use of part-time employees (particularly important in bus operations, which require lots of drivers during rush hour, but far fewer drivers during mid-day) and restrict the types of maintenance work that particular classes of employees are allowed to perform. Public transit agencies also provide generous numbers of sick days, requiring a large "coverage ratio" (i.e., extra drivers available if an unexpectedly high number call in sick).
Administrative inefficiencies and extra layers of management also tend to build up over time. In a study of private contracting worldwide, Gomez-Ibanez concluded that "without the spurs of privatization and competition," public transit agencies decline into a "slovenly lethargy."18
Private bus companies also typically pay lower wages and provide less generous fringe benefits than do public transit authorities. The cuts in fringe benefits are often particularly deep, because fringe benefit rates among public transit agencies are quite high. (The MBTA uses a fringe benefit rate of 39.4 percent.)19 A bus driver at the top of the seniority scale at the MBTA earns a base salary of $40,290. With fringe benefits, the total cost to the MBTA is $57,445 per year. By contrast, a bus driver at the top of the seniority scale at Plymouth and Brockton, a local private bus company, earns $26,520 per year, or 34 percent less than his MBTA counterpart. When fringe benefits are included, the total cost is $34,476. That is $22,969 (or 40 percent) less than the MBTA's costs for a similar position.20 Attempts in Massachusetts to reduce this significant disparity in wages and benefits by contracting privately for bus service have ignited a fierce battle between management and labor.
Governor William Weld, when first elected, vowed to use private contracting as a way to control costs. In 1993, however, the legislature passed the Pacheco Law, designed to inhibit the use of private contracting. The Law has the following major provisions:
· The purchasing agency must establish quantitative performance measures that the contractor must meet in the areas of quality, timeliness and effectiveness.
· Contractors must pay wages "not less than the minimum wage rate that had been paid _ (to) "regular agency employees."21
· Displaced public employees must (assuming they are qualified) be offered jobs with the private contractor.
· The private contractor must offer equal quality at a lower cost, or better quality at an equal or lower cost.
· The private contractor's bid must be compared not to the current in-house cost, but to the cost of public employees working in the most cost-effective manner possible.
· All contracts must be approved by the State Auditor before they can be awarded.
The Law provides the State Auditor with considerable discretion in applying its provisions, including the ability to issue regulations mandating how to develop cost estimates. Moreover, by requiring the Auditor to approve contracts, the Law sets up an adversarial process. By the time an agency has prepared an RFP (Request for Proposals), reviewed proposals, and selected a private company to manage part of its operations, it has decided, correctly or incorrectly, that a particular service should be delivered via private contracting rather than directly by state employees. The stage for conflict is set by requiring another arm of government, which does not have any expertise in the service being provided, to second-guess the contract.
III. The Conflict
In 1995 the MBTA announced that it intended to become a "virtual transportation agency." The MBTA would continue to control key public policy decisions (such as level of service, fare rates, bus route planning, capital improvements), but daily operations would be contracted privately, shrinking the number of MBTA employees from nearly 6,500 to just a few hundred.
This attempt to contract privately for bus operations is the MBTA's first major step toward becoming a virtual transportation agency. To comply with the Pacheco Law the MBTA has taken the following steps:
1. In July 1996, the MBTA submitted a management study of its in-house operations to the Auditor. The study described the MBTA's bus operations, outlined in-house cost savings initiatives, discussed internal constraints to reducing costs further, and compared the MBTA's costs to those of other transit agencies.
2. In August 1996, the MBTA issued an RFP inviting private companies to bid on all or part of its bus operations. The RFP allowed companies to bid on one or more "bundles" of bus service, with a bundle usually consisting of an MBTA bus garage and all the routes run out of that garage. For bidding purposes, the MBTA organized its entire bus operations into five bundles.
3. On December 31, 1996, the MBTA received bids on two of the five bundles— Charlestown/Fellsway and Quincy.
4. On January 2, 1997 (the next working day, as required by law), the MBTA submitted to the Auditor its estimate of the in-house cost of providing bus service for these two bundles.
5. On April 18, 1997, the MBTA submitted the proposed contracts to the State Auditor. The MBTA also submitted a revised in-house cost analysis (the April Submission), to reflect the fact that bids were received on only two of the five bundles.
6. On May 16, 1997, the State Auditor issued a letter (the May Objection Letter) denying the proposed contracts and detailing his reasons for the denial.
7. On May 23, 1997, the MBTA resubmitted the proposed contracts along with a revised in-house cost analysis (the May Submission). The revised analysis incorporated certain of the State Auditor's findings and provided increased detail on why other findings were, in the opinion of the MBTA, incorrect or invalid.
8. On June 20, 1997, the State Auditor issued a letter (the June Objection Letter) finding that the proposed contracts would not save money and again denied the contracts.
9. On June 23, 1997, the MBTA filed suit in Massachusetts Superior Court, asking the Court to find that the Auditor's review violated the law and that the Pacheco Law itself is unconstitutional.
The exchange of letters between the State Auditor and the MBTA General Manager during this time indicate that both sides perceived the analysis not as a dry accounting exercise but as a high stakes political battle. Each side accused the other of withholding and distorting information. As the following excerpts show, the letters drip with contempt and sarcasm, and reveal a climate in which objective analysis is impossible and in which fact-finding takes a back seat to name-calling. The State Auditor, A. Joseph DeNucci, in his letter of May 16, 1997, said,
[The MBTA's proposals] were so deficient as to preclude this office from conducting a full review _ Although an agency's submission might not be free of error, be complete, or be reasonable, these deficiencies should not dominate each aspect of the submission.
MBTA's General Manager, Patrick Moynihan, responded on May 19, 1997, saying,
Even if we were to assume that every issue raised by your staff during its review was valid (notwithstanding the absurdity of such an assumption), the MBTA's proposed contracts would have still saved a minimum of $16,000,000.
The MBTA and the State Auditor agreed that the amount of the winning contract bid for the Charlestown/Fellsway bundle, the larger of the two proposed contracts, was $243.4 million. Yet they disagreed on the MBTA's own cost to run the system as well as the transition costs of outsourcing this operation. Between January and May, the MBTA submitted three different internal cost estimates.22 The second internal estimate corrected some mistakes in the first estimate, and restored management and overhead costs for the three bundles for which MBTA did not receive bids. The third submission corrected some mistakes the Auditor found, and defended certain cost estimates which the Auditor alleged were incorrect, but which the MBTA insisted were accurate.
Before reviewing the specific charges and counter-charges regarding the cost of operating the Charlestown/Fellsway bundle, Table 2 below outlines the scope of the services to be contracted—about one-third of the MBTA's entire bus operations.
Dimensions of the Charlestown/Fellsway Operation
|Indicator||Charlestown/Fellsway||Percent of MBTA Total|
|Peak Bus Requirement||241||32|
The MBTA claims there is an avoidable cost (the cost that can be avoided by privately contracting) for running this bundle of either $262.7 million (January), $261.2 million (April), or $261.0 million (May). The MBTA's estimates of transition costs range from a net gain of $1.9 million (April) and a net loss of $0.6 million (May). The MBTA also estimates that it needs to spend $2.3 million to monitor the contractor over the five-year life of the contract. Taken together, the MBTA claims total five-year savings ranging from $17.5 million (April) to $14.8 million (May). The MBTA's analysis projects savings of 5 percent—far less than other transit agencies report, primarily because the contractor proposed to match the wages of MBTA employees to ensure that the contract met the terms of the Pacheco Law.
In ruling that the MBTA cannot award this contract, the Auditor made the following claims:
A. The MBTA made several mistakes in its in-house cost analysis with regard to operating and transition costs.
B. The MBTA improperly excluded certain costs from the contractor's bid.
C. The MBTA made unreasonable assumptions regarding the number of layoffs that will result, and the associated unemployment compensation, retirement and vacation buy-out costs.
D. The MBTA's accounting system is so deficient that the Auditor's office had insufficient information to make a comprehensive review.
E. The MBTA did not adequately establish that quality of service would be equal or better.
Finally, the Auditor raises a number of other issues, including contractor compliance with environmental laws, the lack of a contingency plan for work stoppages, the value of union concessions, and procurement irregularities.
A. Operating and Transition Cost Errors
In examining the MBTA's documentation of its internal costs, the Auditor notes the following errors:
|Double counting fuel tax savings||$2,520,973|
|Incorrectly projecting annual increases in fuel taxes||$130,242|
|Erroneous building and maintenance charges||$263, 595|
|Incorrectly assuming rental savings||$75,608|
|Incorrect wage inflation rate for executives||$168,202|
SOURCE: A. Joseph DeNucci, Auditor of the Commonwealth, letter to Patrick J. Moynihan, General Manager, MBTA, May 16, 1997 (the May Objection Letter). As the Auditor does not specify the amount for annual increases in fuel taxes, rental savings, and wage inflation, I have calculated these amounts from information contained in the MBTA April and May submissions.
The MBTA acknowledges these mistakes (except for wage inflation for executives, which it disputes yet agreed to change), but notes that even after correcting these errors, the contract would still produce significant savings.
B. Mistaken Contract Costs
The Auditor alleged that the MBTA incorrectly excluded two potential contract costs: incentives for improved service, and liquidated damages should the MBTA terminate the contract early.
The proposed contract allows the contractor to earn incentive payments for superior service, which would include running at least 99.75 percent of all scheduled bus trips. The Auditor noted that the MBTA has already achieved this percentage of trip completion, allowing the contractor to earn an incentive for simply matching the existing level of service. The Auditor concluded that such incentives should be included in the base price of the contract; and he is probably correct. While the MBTA does not routinely achieve 99.75 percent trip completion, the fact that the MBTA met this standard during a recent quarter supports the Auditor's claim that the amount of this incentive, which the Auditor places at $271,515, should be added to the amount of the contract.
The contract allows the MBTA to terminate the agreement at any time for any reason (termination without cause). This clause is frequently found in government contracts and increases the contractor's financial risk due to start-up costs for every major outsourcing contract. Typically, contractors amortize those costs over the life of the contract. Thus, if the MBTA were to terminate the contract without cause after only a few months, the contractor would be unable to recoup its start-up costs. The contract therefore contains a liquidated damages provision to protect the contractor against such an event. The provision requires the MBTA to pay $740,000 if the MBTA terminates the contract without cause during the first year. The amount of liquidated damages declines each year, corresponding to the amount of start-up costs the contractor should be able to recoup, as reflected in the price of the proposed contract submitted by the contractor.
The Auditor seems to have mistakenly concluded that this provision is intended to compensate the contractor for its possible loss of profit if the contract were to be terminated early. In his May Objection Letter the Auditor states, "In the absence of any data to the contrary, the payment for "projected profit" for services not performed is clearly not in the public interest."23 However, the MBTA's submission shows that the contractor projects a profit of $23.2 million over the five-year contract.24 Thus, while the amount for liquidated damages ranges from $740,000 down to $60,000, the intent of this clause is clearly not to ensure that the contractor meets its projected profit goal, as the Auditor claims.
C. Unreasonable Cost Savings Estimates
The Auditor claims that the purported cost savings from this contract rest on a number of unreasonable assumptions. The most important of these is the MBTA's estimate of the number of employees to be laid off, and to what degree these employees will be eligible for compensation under federal law. Other assumptions relate to the number of employees eligible for retirement benefits, and the proper treatment of accumulated vacation time.
Lay-offs, unemployment and 13(c) protection. The MBTA currently employs 628 staff at the Charlestown/Fellsway bundle. The contractor estimates that it will hire up to 559 of those staff. Net savings to MBTA would largely depend upon the contractor's ability to reduce staff costs by reducing the number of employees, and the cost to MBTA of laying off employees. In researching its private contracting options, the MBTA found that private contractors typically hire 70 to 90 percent of the employees of the public transit authority doing the work to be contracted. The MBTA assumes that the contractor will hire 85 percent, or 519, of the 628 affected employees, leaving 109 potentially without jobs. The MBTA estimates that 57 of these employees can be placed elsewhere in the MBTA as vacancies occur, leaving 52 employees who would be laid off.25
Each employee lay-off will cost the MBTA an average of $4,900 in unemployment benefits. However, the MBTA faces a greater potential liability under a federal law commonly known as "13(c)," which is why it must accurately estimate the number of employees to be laid off. Section 13(c) of the Urban Mass Transit Act of 1964 (49 U.S.C. s5333[b]) provides employees of transit authorities that receive federal funds (such as the MBTA) with significant protection, including:
· Preservation of rights, privileges and benefits under existing collective bargaining agreements;
· Continuation of collective bargaining rights;
· Protection against a worsening of their employment position;
· Priority of re-employment for employees who are laid off; and
· Paid training programs.
Section 13(c) "is an especially complex Federal requirement _ administered by two Federal agencies _ which do not always share the same policies or interpretations."26 The Department of Labor, which has primary responsibility for administering this provision, takes the position that,
Section 13(c) agreements should be the product of negotiations between the parties and therefore has not prescribed, for general application, the requisite elements of a Section 13(c) agreement. Thus, there in no single regulation, policy statement or other guidance that summarizes what Section 13(c) requires.27
Instead, when management and labor disagree, the matter is often resolved by arbitration or by the courts.
Under Section 13(c), dismissed or displaced employees can receive a "dismissal" allowance equal to the difference between their old wages and their new wages. For example, a bus driver earning $3,000 a month who was laid off and found a new job at $2,000 a month would be eligible for a dismissal allowance of $1,000 per month. The allowance would continue for the length of time the employee had worked for the transit agency, up to a maximum of six years. Unions have attempted to widen the protection offered to "worsened employees" to include any worsening in fringe benefits, working conditions or hours.
To avoid testing its 13(c) liability, the San Diego rapid transit agency contracted its bus operations in small segments so that it could offer affected drivers the option to fill vacancies elsewhere in the transit agency. However, other agencies (notably Las Vegas and Indianapolis) have taken the opposite approach and have privately contracted most or all of their bus operations in a single swoop.28
The MBTA, in part to minimize its 13(c) liability, developed a procedure it expects will result in layoffs being concentrated among employees with the lowest seniority. Nonetheless, the MBTA and the Auditor have sharply different estimates of this liability. The MBTA estimated its liability at between $2.9 and $4.3 million,29 while the Auditor suggested a minimum liability of $36 million.30 While the MBTA provided considerable analysis to justify its range of liability, the Auditor did not provide a comparable analysis in his objection letters, and did not document how he arrived at his estimate of $36 million.
Retirement Savings. After 10 years of service, MBTA employees become vested in MBTA's retirement program. Once vested, the MBTA is obligated to provide certain retirement benefits. By concentrating layoffs among employees with the least seniority, the MBTA expects to avoid laying off employees who are already vested in the retirement system. Accordingly, the MBTA assumed that it will incur no pension liability during the transition, and that it will ultimately save pension costs, as individuals who would have eventually retired and drawn an MBTA pension go to work for a private contractor instead.
The Auditor noted that many of the MBTA's employees have five or fewer years of service and thus would not be vested at the end of the five-year contract. The Auditor then seemed to conclude that because these employees would not become vested during the term of the contract, the MBTA would not have any pension obligation to them and therefore could not claim any pension savings. This position is correct only if the MBTA plans eventually to bring the work back in-house. Since MBTA management plans to increase the use of private contracting, and no other transit agency has brought bus service back in-house, the Auditor's position is indefensible. Moreover, the Auditor made no attempt to quantify the amount by which he believes the MBTA has overstated savings.
Vacation Buy-Out. The MBTA will have to pay the accumulated vacation leave of the estimated 571 employees who would leave the MBTA because of this contract. The MBTA asserted that this cost should not be included as a transition cost because the vacation leave was earned prior to the start of the contract, and the MBTA will eventually have to pay it regardless of whether this contract is approved. Either employees will quit or retire with accrued vacation time (and get paid for it), or they will use their additional vacation time, and the MBTA will have to pay other employees to perform the work. Although the Auditor takes the position that vacation buy-outs should be included in the transition cost, he did not attempt to estimate the value of this accrued vacation time. Assuming that the average employee has two weeks of vacation time, and that the average weekly wage is $800 (the top of the scale for a bus driver), the total liability would be about $913,000. Even so, the MBTA's position is correct. If the contract is denied, accrued vacation time will remain as a liability on the MBTA's balance sheet. If the contract is approved and the MBTA pays $913,000 in accrued vacation time, it will have less cash but it will no longer carry that liability, and therefore its net financial position will be unchanged.
D. Accounting Deficiencies
Many of the changes the MBTA made from one submission to the next reflect a reallocation of overhead costs. In its April 1997 submission, which included the proposed contracts, the MBTA made two significant sets of changes. First, it reduced the amount of administrative labor allocated to the Charlestown/Fellsway bundle by $5 million. Yet since it received bids on just two bundles, the MBTA could not cut overhead and administrative costs as much as it could have if it were able to contract privately for its entire bus service.
Changes in Overhead Cost Allocation:
January to April
|Item||January Estimate||April Estimate|
|General Administrative Wages||$3,432,924||$2,384,101|
|Bus Operations Administrative Wages||$10,454,380||$8,182,245|
|Fringe Benefits (39.82%)||$5,529,924||$4,207,519|
Second, the MBTA made upward revisions in its cost estimates for materials, tires, diesel fuel, uniforms, supplies, and services, partly offsetting the $5 million loss of savings described above.
Revised Estimates for Cost of Materials:
April and May
|Item||April Estimate||May Estimate|
SOURCE: January Submission, Form 2A; April Submission, Form 2A
In a July 1996 report to the Auditor, the MBTA acknowledged, "It is difficult to be precise regarding expenditures for each transit mode within the Authority—the accounting system has only recently been tailored to compile each transit mode as a distinct cost center."31 In responding to the Auditor's questions regarding why these changes were made, the MBTA noted, "Actual costs for materials, supplies and services are 'mapped' to the appropriate account in the general ledger from the purchasing system. This mapping was imperfect for FY96."32 Finally, in its May submission, the MBTA increased the cost of maintaining "non-revenue vehicles" (everything other than buses) by over 50 percent, from $830,708 to $1,273,209.33
Changes of this magnitude are not unusual for a government agency. Very few government accounting systems are capable of providing an accurate analysis of costs for any given sub-operation. For example, the MBTA needs to know how much it spends on bus tires so that it can properly monitor its current year budget and accurately prepare its budget request for next year. But before the MBTA developed its plan to contract privately for bus operations, it had no real need to know how much it spent on tires in the Charlestown/Fellsway bundle.
In total, the MBTA reduced its claimed savings by $5 million, and then increased its claimed savings by $2.2 million, for a net reduction of $2.8 million. As the MBTA's total five-year cost for running the Charlestown/Fellsway bundle is $287 million, this net change of $2.8 million represents just 1 percent of the total cost and would not appear to support the Auditor's depiction of the MBTA's accounting system as wholly deficient. However, the timing of certain changes gives the impression that the MBTA is being less than forthright about its internal cost allocation methods.
In December, the MBTA learned that it received bids on only 40 percent of the bus system, and therefore needed to carry $5 million more in management overhead related to the Charlestown/Fellsway bundle. Before submitting its April cost comparison, the MBTA reviewed its materials, tires, diesel fuel, uniforms, supplies, and service line items, and concluded that all six of these line items should have had higher amounts allocated to the Charlestown/Fellsway bundle. That increased projected savings by $1.8 million and offset more than one-third of the $5-million loss.
In his May Objection Letter, the Auditor noted several MBTA mistakes worth $469,445. In its May Submission, the MBTA conceded those mistakes but informed the Auditor that it revised its method of allocating non-revenue vehicle repair costs, yielding additional savings of $442,501—neatly offsetting the Auditor's most incontrovertible findings.
It is easy to see how the timing of the MBTA's changes made the Auditor suspicious. While these cost allocation issues are troubling, they do not nullify the MBTA's assertion that the proposed contract would save money. Even if every one of the MBTA's accounting revisions were disqualified, the total amount would only come to $2.2 million. Further, despite the Auditor's point about incentive payments, total savings would still be $12.3 million.
|Item||Amount (in millions)|
|MBTA's estimated savings (May analysis)||$14.8|
|Less: Deny all MBTA cost allocation revisions||(2.2)|
|Less: Incentive payments||(0.3)|
To deny the contract, the Auditor must cast more doubt on the MBTA's analysis. He did this by examining how the MBTA treats repairs made in its Everett heavy maintenance facility. Each bus garage has a staff of mechanics and the facilities to handle routine maintenance and a certain amount of repair work. In addition, the MBTA operates a centralized heavy repair facility (the Everett facility). The Everett operation rebuilds engines and transmissions, repairs damaged frames, and does major electrical systems work. Over the five-year term of the contract, the MBTA estimated it would spend $74 million to operate this facility. The Charlestown/Fellsway contract requires the contractor to purchase repair services from the Everett facility for the first two years of the contract. After that, the contractor can continue to purchase repairs from Everett or make other arrangements.
The Auditor questioned the MBTA's estimate that it would avoid $23.8 million in Everett costs during the five-year contract term. First, the Auditor asked how the MBTA could be sure that it would continue to get revenue from the contractor during years 3 to 5. If the contractor decided not to use the Everett facility during those years, the MBTA would either have to lay staff off (raising additional 13[c] issues) or watch its projected savings evaporate. Second, the Auditor correctly notes,
There is a lack of reasonable assurance that the level of heavy maintenance and repairs to be paid for by the contractors during the required first two service years will approximate the MBTA's estimate thereof. Because the MBTA does not maintain bus repairs by garage, it based its projected cost savings on an estimate of how much of the total $14.9 million budgeted for the Everett facility during the fiscal year 1997 pertains to the Charlestown/Fellsway bundle.34
The Auditor did not offer a specific critique of the MBTA's cost allocation methodology nor an alternative methodology, even though this is a critical issue in determining the MBTA's in-house cost. The MBTA allocates Everett costs to each bundle according to the percentage of the buses each garage needs to operate35—a reasonable methodology. This results in 32.21 percent of the total cost of the Everett facility being allocated to the Charlestown/Fellsway bundle. If a particular bus garage has 32 percent of the bus fleet, it might well result in that garage using 32 percent of total heavy repair services.
There are other ways, equally reasonable, to allocate heavy repair costs. One obvious method is by total miles. The Charlestown/Fellsway bundle accounts for 28.73 percent of total bus miles.36 If allocated according to this percentage, the MBTA's avoidable costs (and thus total savings) would decline by $2.5 million. Another methodology is to use the number of bus breakdowns. The Charlestown/Fellsway buses are relatively new, break down less frequently, and therefore place less demand on the Everett maintenance facility. In 1997, the Charlestown/Fellsway buses suffered 487 breakdowns, or 26.4 percent of total MBTA bus breakdowns.37 If the MBTA were to use this methodology, total projected savings from this contract would decline still further. Regardless of which methodology is used, the contract would still save money, requiring the Auditor to raise additional objections.
E. Quality Concerns
The privatization law sets up two quality standards. First, the contracting agency must prepare measurable quality standards before it releases its RFP. Second, after receiving bids and negotiating a contract, the contracting agency must certify that the contractor will provide services "that equal or exceed the quality of services which could be provided by regular agency employees."
Performance measures commonly used to evaluate the quality of bus service include the average distance between failures, percentage of scheduled trips run, accidents per million miles, on-time performance, passenger complaints per 100,000 boardings, and cleanliness standards such as daily exterior washings and interior cleanings.
Almost all transit agencies use at least some of these standards, and most public transit agencies put additional quality measures into place when they begin to contract privately for bus service. In a review done for the Texas Public Policy Foundation, transportation consultants Jean Love and Wendell Cox concluded that the quality standards set for most private contracts "routinely exceed those standards previously—and often concurrently—set for service provided by the public authority."38
Public agencies set higher quality standards for private contractors for several reasons:
· Such standards (and the financial penalties for failing to meet them) help deter inexperienced or unqualified firms from bidding;
· The standards provide important leverage to public transit managers, who remain ultimately responsible for service quality even if they are no longer directly delivering the service; and
· Human nature being what it is, it is easier to tighten quality standards when you can hold someone else responsible for meeting those standards.
Injecting competition into any service tends to improve quality. If a monopoly agency (either public or private) lets quality diminish, it is unlikely to suffer any adverse effects for a very long time. Ridership will decline, but the operation will continue as long as public money is available to subsidize it. By contrast, a company that wins a competitively awarded contract has a strong incentive to meet quality standards. Failing to meet those standards can result in financial penalties, and may ultimately result in contract revocation or non-renewal.
MBTA followed industry standards. In attempting to ensure that quality remains at least equal to current standards, the MBTA followed what is becoming standard industry procedure:
First, the MBTA has negotiated a contract that incorporates more stringent quality measures. Because many of these performance measures are new, the MBTA and the contractor must agree to establish a baseline before actual performance levels can be established, but this is not uncommon.
Second, the MBTA has established financial incentives and penalties. If the contractor meets all performance measures, it can earn an extra $2.3 million. If the contractor fails to meet these incentives, it is subject to a maximum penalty of $2.5 million.39 As the contractor projects a baseline profit of $23.2 million from this contract, meeting quality measures can mean the difference between a profit of $25.5 million and $20.7 million, giving the contractor a significant financial incentive to improve quality.
Finally, the MBTA has selected an experienced contractor. ATC/Vancom is providing bus service to a number of other public transit agencies and has had several of its contracts renewed by these agencies. The company has experienced management and adequate financial strength to undertake a contract of this magnitude.
Nonetheless, the Auditor has concluded, "The MBTA does not know whether the proposed contractors can meet acceptable quality service levels, let alone exceed the quality of service that is currently provided by regular MBTA employees."40 It is important to note that the Auditor has not provided evidence that the quality of service likely to be provided by the contractor is inferior. Instead, the Auditor determined that the MBTA did not meet the Pacheco Law's requirements for service quality.
The Auditor's criticisms. The Auditor's criticism of the MBTA's efforts to ensure quality service is the weakest part of his review. These are his specific criticisms:
(1) Insufficient benchmarks. The MBTA proposed to create new quality measures and hold the contractor responsible for meeting those measures. The Auditor faults the MBTA for not having developed those measures earlier and submitting them as part of its July 1996 management study.
The MBTA should be commended, not criticized, for its attempt to write additional quality indicators into the proposed contract. This is standard industry practice and helps give the public transit authority important leverage over the contractor, especially when the quality standards are combined with financial incentives/penalties, as the MBTA has done.
(2) Failure to establish maximum in-house level of service. The Auditor noted the following:
(E)xisting standards do not correlate to Chapter 296 quality requirements which could be provided by regular agency employees. Accordingly, we believe that the need to establish appropriate performance factors for each pertinent performance attribute based on the quality of service which could be provided by regular agency employees is a relevant fact for the MBTA to clarify.41
Estimating the quality of service that "could be provided by regular agency employees" is virtually impossible. For example, sometimes the MBTA is unable to run scheduled bus trips because too many bus drivers call in sick. The MBTA could develop and implement new policies to reduce sick leave utilization. However, it is impossible to determine what effect this would have on the percentage of trips run until such a policy were actually implemented.
(3) Lack of thorough reference checks. Finally, the Auditor noted that the MBTA did much of its reference checking on the contractor after it submitted the proposed contract to the Auditor on April 18, 1997. The Auditor concluded that this calls "into serious question the integrity and completeness of the MBTA's quality comparison."42
Conducting site visits and reference checks after negotiating contracts is a flaw in the MBTA's procurement process. However, the MBTA knew that the proposed contractor is experienced in the field, and knew that other transit agencies which privately contract for bus service have reported receiving equal or better quality service. The reference checks, although performed late, were uniformly positive.
F. Other Issues
Compliance with environmental laws. Section 54(7)(iv) of Chapter 7 requires the contracting agency to certify that its proposed contractor "have no adjudicated record of substantial or repeated willful noncompliance with any relevant federal or state regulatory statute." In his May Objection Letter the Auditor noted that the MBTA did not submit adequate documentation regarding the contractor's compliance with environmental laws. The MBTA responded that the contractor had provided assurances that it met this standard and that the state Department of Environmental Protection reported that the contractor did not have a record of substantial or repeated violation of environmental laws.
Lack of contingency plans. The Auditor noted that MBTA employees are enjoined from striking, but that employees of the contractor are not so enjoined, and concluded that "it is essential that contingency plans be prepared setting forth the measures that the contractor plans to institute in the event of employee work stoppages." In his review of private contracting for bus operations, Wendell Cox reported that only five days of service have been lost as a result of contractor default in the United States during the last decade.43 Given this level of risk, it seems less than essential that the contractor develop contingency plans in the event of a strike.
Procurement irregularities. After receiving the bid, the MBTA negotiated some changes in both the contract and the price. The most important change was to advance the contractor 75 percent of the monthly contract cost, or about $3 million per month. In exchange, the contractor reduced its bid by $1.9 million. The Auditor opines that "any major revisions to the plans and specifications requested by a successful bidder during contract negotiations after bid opening must be denied or the procurement must begin anew."44 This view is not widely held and is not in the public interest. Many states conduct a negotiating session known as "best and final offer." During such negotiations, the purchasing agency and the contractor discuss any changes to the contract that might result in lower cost and better value to the taxpayer. In this case, it looks like the MBTA and the contractor realized that because the MBTA can borrow at tax-exempt rates, it would be more advantageous to the taxpayers for the MBTA to borrow the money and advance it to the contractor. Surprisingly, the Auditor did not seek to estimate how much interest the MBTA would pay and subtract that amount from the total contract savings. The MBTA's interest costs will of course depend on actual interest rates over the next five years, but at current rates the MBTA should expect to pay $1 million. While it would be legitimate to add this amount to the cost of the contract, the Auditor has overreacted by claiming that this change invalidates the entire contract.
Union concessions. In an attempt to prevent the MBTA from privately contracting any part of bus operations, the MBTA's unions developed a cost savings package and submitted it to MBTA management. The specific cost savings were to go into effect only if they were needed to reduce the in-house cost below the lowest bid by a private contractor. The unions value their concessions at $29.5 million for both bundles.45 MBTA management disagrees with many of the unions' proposed cost savings, and puts the value of proposed savings for the larger Charlestown/Fellsway bundle at $3 million.46 Thus, if the MBTA is correct, the private bid still represents the lowest value. However, if the unions are correct (or even half correct), then their package of cost savings results in the lowest bid, and the work should be kept in-house. The Auditor made no attempt to analyze this issue. Instead, he warned that if the MBTA succeeds in overcoming his existing objections, he "would be required to place a value on the submitted concessions, and to further determine whether the value of those concessions would negate the identified amount of savings."47
The Auditor has raised numerous objections to this proposed contract. In certain cases, such as incorrectly calculating savings from fuel taxes and other items, he is clearly correct. In other cases, his arguments are questionable. Finally, he made several objections that detracted from the overall credibility of his position (e.g., quality issues, lack of contingency plans, compliance with environmental laws, procurement irregularities). Yet his most important objection remains the reliability and accuracy of the MBTA's accounting system.
The MBTA claims that over the five-year contract period, it would save (i.e., avoidable cost) $261 million in operating costs for the Charlestown/Fellsway bundle alone. Table 5 shows the major categories of expense:
Avoidable Cost for Charlestown/Fellsway
|Expense Item||Five-Year Cost
|Percent of Total|
|Direct Wages and Fringe Benefits||$195.1||74.8|
|Materials, Supplies, and Services||$20.9||8.0|
|Other Direct Costs||$9.7||3.7|
|Everett Heavy Repair||$23.8||9.1|
|Non-Revenue Vehicle Repair||$1.3||0.5|
|Indirect Departmental Management||$10.2||3.9|
The Auditor did not object to two categories of expense: direct wages and fringe benefits, and indirect departmental management. Together these two items account for 78.7 percent of the total avoidable cost. Moreover, while the Auditor has pointed out some minor mistakes or invalid assumptions in "other direct costs," he did no say he found the MBTA's underlying accounting system to be unreliable in this area. When combined with the two line items mentioned above, it appears the Auditor accepted 82.4 percent of the total cost savings as based on reliable accounting.
Even so, the Auditor questions the MBTA's accounting and cost allocation system in the remaining three areas: materials, supplies, and services; Everett heavy maintenance; and non-revenue vehicle repair.
A. Materials, supplies and services
The MBTA's documentation in this area has been problematic. For example, in April 1997 the MBTA told the Auditor that actual materials costs for fiscal year 1996 were $237,869, but that it was budgeting almost three times as much ($671,149) in fiscal year 1997. Other line items were also dramatically revised, some up, some down. However, while the individual line items often changed dramatically, the overall total did not. In fiscal year 1996, the MBTA reported spending a total of $3.9 million on material and supplies for the Charlestown/Fellsway bundle. In April 1997, the MBTA reported it was budgeting $3.6 million for this bundle in fiscal year 1997, and in May, it reported that it was increasing the budgeted amount back up to $3.9 million (see Table 4). While these changes are unfortunate, it is important to remember that the maximum possible difference between the three estimates is just $0.3 million, hardly enough to justify denying the award of the contract on the basis of an unreliable accounting system. But with the MBTA itself unsure of exactly what its costs are in this area, it would be reasonable for the Auditor to require that the MBTA use its lowest estimate (contained in its December 31, 1996 in-house estimate). This would reduce total savings from this contract by $1.8 million.
B. Everett heavy maintenance and non-revenue vehicle repair
The Auditor was on solid ground when he questioned the methods used to allocate a percentage of the cost of running the Everett heavy repair facility and the non-revenue vehicle repair facility. As we have seen, the MBTA has apparently used a cost allocation methodology for Everett that results in the highest possible savings. A middle-of-the-road methodology would reduce savings by $2.5 million. In addition, the Auditor can reasonably deny the MBTA's 52 percent increase in its cost estimate for non-revenue vehicle repair (as it was made without adequate explanation), reducing savings by another $0.4 million.
Not one of these issues is serious enough to justify denying the contract. At most, taking the Auditor's objections into full consideration, the MBTA's projected five-year operating cost savings would be reduced by $4.7 million, or 1.6 percent. Cost allocation methodologies are necessarily imperfect. Indeed, the uncertainty in the MBTA's accounting system is an argument in favor of private contracting. If the MBTA were to contract privately for a portion of bus operations, it would finally know exactly what this service costs, and it could use that knowledge to benchmark and manage the cost of the 60 percent of bus operations that would remain in-house.
The Auditor's next justification for denying the contract was his determination that the contract would not save money. Table 6 summarizes the Auditor's objections to the contract. Where the evidence supports the Auditor's finding, the chart deducts the appropriate amount from the MBTA's claimed savings. Where the evidence does not support the Auditor's finding, the table contains a zero in the amount column. Further, where the Auditor did not attempt to quantify the amount of his dispute with the MBTA, I have supplied my own estimate. The starting point for this table is the MBTA's final cost analysis (May 24, 1997). That analysis assumes cost savings of $14.8 million and incorporates several of the Auditor's objections (i.e., fuel tax, rent, building and maintenance). Accordingly, Table 6 displays only the unresolved issues.
Auditor's Objections to the MBTA's Cost Analysis
|Materials, Supplies, and Services||$1.8|
|Retirement and Post-Employment Benefits||$0.0|
|Everett Heavy Repair||$2.5|
|Interest on Advance Payments||$1.0|
Although this analysis gives the Auditor the benefit of the doubt, the contract still saves $8.8 million. This amount should be further reduced by the MBTA's 13(c) liability, which the MBTA estimates will range from $2.9 to $4.3 million. Using the mid-point of this range ($3.6 million), ultimate savings would be reduced to $5.2 million. Although the MBTA estimated that it could reduce its in-house costs by $3 million if it accepted the concessions made by the union, private contracting is still the most cost-effective option.
Virtually every transit authority that contracts privately for bus service achieves savings of 20 to 30 percent. Yet the MBTA, in its own analysis, estimated savings of just 5 percent. The Auditor identified several problematic elements of the MBTA's analysis that could reduce total savings to $5.2 million or 1.8 percent (this does not include his more improbable assertions).
The differences between the Auditor's estimate, the MBTA's estimate, and the experience of other transit agencies raises three issues: (A) Why are potential savings from contracting at the MBTA so much lower than elsewhere in the nation? (B) How can we be sure which party is right or if either party is correct? (C) What does this conflict teach us about the Pacheco Law?
A. Why MBTA savings are lower
Based on the experience of other agencies, the MBTA should be able to save 20 to 30 percent ($57 to $86 million) over five years by contracting privately. Instead, savings were estimated to range from $5.2 million to $14.8 million. The two main reasons for the discrepancy are the MBTA's in-house cost accounting method used to determine costs, and the Pacheco Law provision that requires contractors to provide substantially similar wages.
The MBTA's internal costs can be divided into two main elements: direct costs and indirect costs. Direct costs are such things as the wages and benefits of bus drivers and mechanics, and all the materials and supplies consumed by a bus operation. Indirect costs are primarily MBTA management and support services, such as MBTA senior management time, the law office, human resources, the treasury department, and other central services. In a cost allocation scheme, a transit agency might allocate one-third of its senior management cost to bus operations on the assumption that they spend that proportion of their time on buses, with the rest being devoted to subway, trolleys, commuter rail, etc. The Federal Transit Authority (FTA) recommends that transit authorities include all such indirect costs when they attempt to determine their internal costs. The FTA assumes that such costs will, in the long-term, decline if a transit agency privately contracts for a significant amount of its operations. Many of the 20 to 30 percent savings figures are based on this so-called full cost allocation method.
In Massachusetts, the Auditor excludes certain indirect costs from cost comparisons. His rationale is that many indirect costs will continue over the life of the contract and are therefore not "avoidable." He used this rationale when he excluded a portion of the rent for MBTA headquarters, noting that the MBTA would still have to make the same rental payments even though it was employing a few less staff at headquarters. In the short term, his position is accurate. In the long term, however, if the MBTA expands its use of private contracting it will need significantly less headquarters space. Thus it is reasonable to assume that the MBTA would take less space when its lease comes up for renewal. Also, since the MBTA excluded its short-term debt (which most other transit agencies do not have to pay), it did not include $19.2 million in in-house costs. If the MBTA had followed the FTA's methodology, most of these costs would be included in its in-house cost, resulting in much higher potential savings.
The Pacheco Law requires contractors to pay wages that are not lower than the lesser of (a) the minimum wage rate that has been paid for those positions for which the duties are substantially similar to the duties performed by regular employees of the MBTA, or (b) the average private sector wage paid for comparable positions.
Labor is the primary cost of bus operations, making up at least 80 percent of the total cost. The potential savings from private contracting are greatly diminished if the private contractor is locked into the same wage rates and fringe benefits as the public agency. One study of this issue concluded that 62 percent of the private sector's cost advantage was due to lower wages and fringe benefits.48 It is not unusual for a private bus company to pay $5 per hour less than a public transit agency. As this contract involves about 590 employees, a cut of $5 per hour would reduce wages by $6.1 million annually, or about $30.5 million over five years. Add in payroll taxes that vary with wages, and the total comes to about $35 million in savings over five years.
If full-cost accounting is applied and if the Pacheco Law did not exist, the MBTA would have received bids based on lower wages, achieved additional savings of about $54.2 million, and reached the 20 percent threshold that other transit agencies report as minimum savings.
B. How can we be sure who is right?
You don't have to believe in the old adage, "figures lie and liars figure" to be confused about whether the MBTA was correct when it claimed that it would save $14.8 million, or the Auditor was right when he claimed that there would be no savings. The two main areas in which the Auditor and the MBTA differ involve cost allocation methodologies, and assumptions about the future. I have already discussed the MBTA's cost allocation methodologies and concluded that, at most, the MBTA overstated savings by $4.7 million. Now let us assess the reasonableness of the MBTA's assumptions about future costs.
The MBTA acknowledges that cost savings ultimately depend on unknown and unknowable factors, such as the number of bus drivers to be laid off, and how an arbitrator or judge will apply 13(c) protections. There are many assumptions that the Auditor has not questioned, but that can determine the outcome of the review. For example, the MBTA assumes that wages will rise 3 percent per year over the life of the contract, for a total increase of 12.2 percent. Yet neither the MBTA nor the Auditor can predict labor market conditions over the next five years. Moreover, if MBTA management cannot agree on a wage increase with its unions, the matter will be settled via arbitration. In the past, arbitrators have shown a willingness to make significantly higher awards than what the MBTA offers. This is of particular concern given today's tight labor market. Unemployment is at or near a 20-year low, and wages are beginning to rise more quickly than they have in several years. If this trend continues, the MBTA's predictions of a 3 percent wage increase may prove unrealistic and in-house costs may be significantly higher. Assuming a 4 percent average annual increase in wages (instead of 3 percent) would increase total in-house costs by nearly $4 million. This type of sensitivity analysis was not performed by either the MBTA or the Auditor, but can determine the final result.
C. Pacheco Law Flaws
The Pacheco Law has slowed the pace of competitive contracting in Massachusetts to a crawl. In the three years before the Law was passed, the Weld administration issued 36 private contracts, but in the four years since the Law was passed, just 7 contracts have been submitted to the Auditor. From a public policy perspective, the Pacheco Law is poorly constructed. Its more important flaws include the following:
Avoidable cost standard. The Auditor requires that the full cost of a private contract be compared to the avoidable in-house cost, even though the FTA recommends that transit agencies determine in-house costs on a full-cost rather than an avoidable cost basis. The FTA reasons that if the full cost of the first, small-scale private contract is compared to a less-than-full in-house cost, then private contracting will be unable to gain a foothold. If private contracting is truly less expensive than the full in-house cost, and its use is gradually expanded, then over time the contracting agency will indeed be able to reduce management, rent and other costs that at first appear unavoidable. These savings will never be realized if the first attempt to introduce private contracting is denied because the private contract is compared to a less-than-full in-house cost.
Quality standards. It takes time (and money) to develop enough data to establish reliable benchmarks. The Auditor's insistence that the MBTA develop such measures could, in theory, indefinitely delay the MBTA's ability to contract privately for bus service. Other transit agencies routinely develop new quality measures when they begin to contract privately, and there is no reason that bus riders in Massachusetts should be deprived of whatever improvements in service those measures can effect. Moreover, if the Auditor can disagree with the contracting agency over a relatively simple service such as bus operations, the scope for potential disagreement over how to measure the quality of complex human and social services is far greater. Finally, the Auditor's emphasis that the contractor's proposed service be measured not against the current level of service but the level of service that could be provided creates an impossible hurdle.
Raising new issues and keeping issues in reserve. The MBTA claimed that the Auditor raised new issues in his second objection letter. The Auditor claimed that he was justified in raising those issues because the MBTA provided new information in its second submission. In addition, the Auditor warned the MBTA that if it ever succeeds in meeting his earlier objections, he will raise the issue of the unions' concessions. Clearly, this combination of raising new issues and keeping issues in reserve offers the possibility of endless delay.
The Auditor does not have to say what he thinks are the MBTA's internal costs. This allows him to simply raise objections (e.g., What if more MBTA employees are laid off? What if the 13[c] ruling is less favorable than MBTA management expects? What if the contractor does not use the Everett repair shop after the second year?) rather than to take an affirmative position on these issues. It is as if the Auditor sees his role as that of a defense lawyer, trying to plant a reasonable doubt, rather than come to an independent and defensible cost comparison. Furthermore, the Auditor himself acts as both judge and jury, determining whether the doubts he has raised are sufficient to justify denying the contract.
This struggle between the Auditor and the MBTA demonstrates that Massachusetts needs a fundamentally different strategy for determining whether a particular private contract is likely to save money. In most states, this task is left to the judgment of the agency entrusted with providing the service involved. Massachusetts has attempted to replace this judgment with a cost accounting exercise. But we have seen that no matter how much analysis is done, in the end certain issues come down to personal judgment. The Auditor's ability to simply criticize and cast doubt on a contract without providing his own cost estimates gives him an arbitrary power to approve or deny contracts.
If the state legislature wants an independent, objective review of contracts, it must revise the Pacheco Law as follows:
1. The scope of the review should be narrowed to cost. Quality is very difficult to measure and almost impossible to predict in advance. The authority given the Auditor to claim that certain aspects of a contract are not in the public interest, and to deny the contract on that basis, should be eliminated.
2. The private contract should be compared to existing in-house costs, not the cost that "could be" achieved if the state agency worked in the most efficient manner possible.
3. Assumptions about wage increases, interest rates, and other future variables should be identical to those assumptions the state uses to forecast tax collections. This would provide a consistent set of assumptions, and remove the opportunity for either the Auditor or the contracting agency to determine the outcome of the review by manipulating important underlying assumptions.
4. Agencies should be required to publish cost allocation plans as part of their annual budget requests, regardless of whether they plan to contract privately. Doing so would likely remove the temptation to choose a cost allocation methodology based on whether it would help the contracting agency "prove" that the contract will save money. For example, if the MBTA had to allocate maintenance costs to all five bundles well before it put out an RFP, there would have been no incentive to choose a methodology that maximized the cost of the Charlestown/Fellsway bundle because they did not know in advance the bundles for which they would receive bids.
5. The Auditor's review of an agency's accounting system and cost allocation plan should be completed before an agency issues an RFP. If the Auditor disagrees with an agency's cost allocation plan, he should quantify the amount of his disagreement, and specify where the amount in question should be allocated. As long as the Auditor must account for 100 percent of an agency's budget, there is no incentive for the Auditor to seek to skew the cost allocation. For example, if the Auditor reduces the heavy maintenance amount allocated to the Charlestown/Fellsway bundle, he has to increase the amount allocated to another bundle(s). While this may make the Charlestown/Fellsway bundle a harder operation to contract privately, it does make other bundles easier to contract out. Since the review must be conducted before an RFP is issued, there is no reason for the Auditor to conduct anything other than a strictly objective review.
6. The Auditor's back-end review should be limited to clear mistakes. It is perfectly appropriate for the Auditor to identify clear-cut mistakes, such as double-counting fuel tax savings. But he should not be allowed to raise generic doubts about an agency's accounting system or cost allocation methodology.
7. The Auditor must come to a firm conclusion. If he finds mistakes, he must quantify them, factor them into the analysis, and offer his own defensible estimates of both the in-house cost and the contract cost.
As this paper was being written, the MBTA was attempting to maximize bus ridership within its existing budget by adding service to its most heavily used routes and by reducing service on lightly used routes. Communities that would lose service have protested the plan intensely. At a public hearing on December 9, 1997, public officials representing the South End, Roxbury, and Jamaica Plain demanded that the MBTA not cut service on three routes in those neighborhoods; and the crowd shouted "Add! Add! Add!" as it called for more bus service, not less.
The estimated savings from contracting privately for the Charlestown/Fellsway bundle would allow the MBTA to maintain the existing level of service on the three routes in question.
1. "Bus Service Delivery System. Final report to the Massachusetts Bay Transportation Authority," Comsis. Prepared in association with Howard Stein/Hudson Associates and John T. Doolittle Associates, Inc., 1993.
2. Jose A. Gomez-Ibanez, "Big City Transit Ridership, Deficits and Politics: Avoiding Reality in Boston," Taubman Center for State and Local Government, Kennedy School of Government, Harvard University, 1994, p. 12.
3. The 1965 and 1991 figures are from Gomez-Ibanez, p. 2.
4. Ibid., Table 7: Transit Projection for 1020 under Different Scenarios.
5. In 1990, MBTA ridership was 660,000 linked trips. Assuming that most riders make two trips per day, this yields an estimate of 330,000 riders. Using the same assumptions to project the 2010 deficit at $1.62 billion, Gomez-Ibanez estimates that ridership will grow 8 percent between 1990 and 2010, producing an estimate of 356,000. If both his deficit and ridership numbers prove accurate, the annual subsidy per rider will come to $4,550.
6. Gomez-Ibanez. p. 8.
7. "Contacting for the Operation and Maintenance of Fixed-Route Bus Service. Charlestown/Fellsway Bundle," Vol. 1, MBTA, May 23, 1997, p. F-19. Hereinafter called the MBTA's "May Submission."
8. The two contracts are very similar, and the Auditor's objections to each contract are almost identical. This paper will focus on a single proposed contract for bus service originating from the Charlestown/Fellsway garage, by far the larger of the two contracts.
9. A. Joseph DeNucci, Auditor of the Commonwealth, letter to Patrick J. Moynihan, General Manager, MBTA, June 20, 1997, p. 2. Hereinafter called the "June Objection Letter."
10. Civil Action No. 97-2547, MBTA v. Auditor of the Commonwealth, First Amended Complaint, June 23, 1997, p. 14.
11. Gomez-Ibanez, p. 15.
12. Fiscal Year 1989 Budget Book, MBTA, p. 4-12.
13. Gomez-Ibanez, Table 5: Components of the MBTA's Deficit Increase, 1970-1990.
14. Jean Love and Wendell Cox, "Competitive Contracting of Transit Services," the Texas Public Policy Foundation and the Reason Foundation, 1993, p. 3.
15. Wendell Cox, "Competitive Contracting in Public Transit: A Review of the Experience," Supplemental Report to the Legislative Transportation Committee, State of Washington, prepared by Cox, Hornung, Lahn, Mundle, and Prestrud, February 1996, p. 3.
16. Wendell Cox and Nick Newton, "Competitive Contracting: The International Revolution and Implications for New York," paper presented at a New York City conference, "Getting from Here to There: A Transportation Plan for New York in the 21st Century," June 18, 1996, p. 15.
17. James McLaughlin, Deputy Executive Officer, Bus System Improvement Planning, Los Angeles County Metropolitan Transportation Authority, paper presented at a New York City conference, "Getting from Here to There: A Transportation Plan for New York in the 21st Century," June 18, 1996, p. 1.
18. Jose A. Gomez-Ibanez and John R. Meyer, "The Political Economy of Transport Privatization: Successes, Failures and Lessons from Developed and Developing Countries," Final Report prepared for the U.S. Department of Transportation under the University Transportation Center, Region One, 1992, p. 4-31.
19. MBTA May Submission, p. D-14.
20. Quincy Patriot Ledger, June 1, 1996, p. 28.
21. Contractors have the option to pay the "average private sector wage for comparable positions," provided that such a rate has been established by the Executive Office for Administration and Finance. Presumably the Auditor must agree that the positions are in fact comparable. Guidelines for Implementing the Commonwealth's Privatization Law, Office of the State Auditor, March 1994, p. 5.
22. The law requires the contracting agency to submit two separate estimates. The estimate of in-house costs is due the day after bids are received. If the contracting agency decides to award a contract, it must next submit the proposed contract, along with a comparison of the in-house cost to the contract cost. The MBTA submitted its in-house cost analysis in January (hereinafter the "January Submission") and the cost comparison in April (hereinafter the "April Submission"). In May, the MBTA submitted a revised cost comparison (the May Submission) in an attempt to satisfy the Auditor's objections.
23. May Objection Letter, p. 15.
24. May Submission, p. E-72.
25. May Submission, p. E-6.
26. Transit Cooperative Research Program, Legal Research Digest, Number 4 (June 1995), p. 1.
27. Ibid., p. 8.
28. A transit agency's 13(c) liability is determined primarily by its contract with its unionized workforce. The fact that other transit agencies have been able to privately contract large chunks of their bus service without incurring massive 13(c) liability does not necessarily mean that the MBTA can do the same.
29. May Submission, p. E-15.
30. June Objection Letter, p. 14.
31. "Management Study: Outsourcing of the Bus Operations and Maintenance Functions," MBTA, July 31, 1996, p. 9.
32. May Objection Letter, p. 6.
33. May Submission, Form 2A.
34. May Objection Letter, pp. 4-5.
35. May Submission, p. E-41.
36. May Submission, p. E-19.
37. MBTA response to Pioneer Institute Freedom of Information Act request.
38. Love and Cox, p. 11.
39. May Submission, Section 3.0, Exhibit I, p. 12.
40. June Objection Letter, p. 14.
41. June Objection Letter, p. 17.
42. May Objection Letter, p. 11.
43. Love and Cox, p. 10.
44. June Objection Letter, p. 19.
45. May Submission, p. F-2.
46. May Submission, p. F-6.
47. June Objection Letter, p. 22.
48. Roger F. Teal, "Issues Raised By Competitive Contracting of Bus Transit Service in the USA," Transportation Planning and Technology, Vol. 15 (1991), p. 402.