FLORIDA:Consultants nix privatization

“The taxable privatization scenarios cut back on needed turnpike widening and expansion projects resulting in high levels of congestion occurring along up to one-third of the system by the year 2020, thereby leaving to FDOT and local jurisdictions the responsibility to fund needed road system improvements that can no longer be funded by turnpike revenues. The taxable privatization scenarios also cut many technology-related projects that are intended to improve patron convenience, operating efficiency and safety (ITS) due to the need to preserve a reasonable rate of return on equity...” the consultants conclude. (p216,7)

Privatization is assumed by IMG to be a commercial operation requiring 12% rate of return on investment, a small reduction of operation and maintenance costs below those of the state, and tolls increasing at ten year intervals with inflation, and an emphasis on gaining non-toll revenues.

The consultants favored strategy is called Enhancement and involves implementation of improved management practices within the department’s Florida Toll Enterprise (FTE).

“The Enhancement option permits the turnpike to finance and implement a significantly greater system expansion program than either the (as-is Option A) Base Case (27% greater) or (Option B-1) the taxable privatization options due to the combined effects of (1) a less conservative debt service coverage requirement, (2) more aggressive pursuit of non-toll revenues, and (3) more efficient administration and operations.” (translated from the consultantese)

Each strategy is assumed to generate approximately the same total revenue spread over 20 years and discounted to present value at a 5% discount rate: $9.5b. So each starts with about the same revenue ‘pot.’ The investor-owned turnpike would pay $1.3b of that in federal and state taxes and $1.6b for defeasement of the debt, $470m in dividends and $160m in royalties to the state. That would leave $6.2b for debt service, capital spending, and operations and maintenance. In state hands the turnpike would have no dividends, no taxes and no royalties to pay, so the consultants suggest the whole $9.5b could be deployed for capital and operating costs and debt service.

All the financial modeling assumes across-the-board toll increases of 25% in 2006 and again in 2016. A major argument for privatization is that, structured correctly, it can generate substantially larger toll revenues through charging more what the traffic will bear, rather than what the politicians in charge of state and county toll facilities find electorally expedient. And Florida’s toll rates are low even by the standards of other state toll agencies. The ‘same-pot’ assumption deals privatization options a heavy blow right off!

IMG calculates that the No-Change arrangement (Option A) – in which the Turpike District and Office of Toll Ops remain separate – will allow funding of a $5.1b capital program over 20 years, while preserving a net-debt service coverage of 1.7. This supports an extra 826 lane-km (513 lane-mi), a 30% addition to the capacity of the Turnpike system, 12 new interchanges, 3 ramp projects and other improvements. Of the 826 lane-km only 70 lane-km (44 lane-mi) or $200m worth are new tollroad, the rest, $1.6b worth being widenings. This is estimated to keep all but 6%, about the present number, operating at LOS-F or better.

Option B, privatization, would generate a purchase payment for the state of $2.2b or more, IMG estimates. But a large slice of this would simultaneously go toward defeasing the $1.9b Turnpike debt. The other financial benefit to the state, IMG says, would be $177m in state taxes paid over the 20-year period. Full privatization to a for-profit corporation supports a capital program of only $2.2b, 307 lane-km (192 lane-mi) an 11% increase. Partial privatization, or sale to a tax-exempt not-for-profit would support a $4.1b capital program extending the system by 707 lane-km (442 lane-mi) or 26%. Full privatization would leave 32% of the turnpike system at LOS-F, while partial privatization to a non-profit would leave 12% of the system at LOS-F in 2020, versus 6% with the consultants’ preferred Florida Turnpike Enterprise within FDOT.

“(The) Turnpike capital program must be curtailed by up to $2.9b to provide the financial capability to support a reasonable rate of return on equity for a taxable private sector (toll) entity,” say IMG. (p141) This is their decisive argument against privatization.

The consultants are not so negative about contracting-out of operations, however, saying this can support a $6.5b program, adding close to 1,000 lane-km (620 lane-mi) or 36% to the system.

They do list ways an entrepreneurial private operator could enhance turnpike revenues, but add that the Turnpike Enterprise proposed within the FDOT might also develop them. The extra revenues might come from advertising on turnpike facilities, toll tickets, transponders, branding programs, congestion pricing, toll buy-in or HOT lanes on FL interstates, leasing of enhanced real estate for hotels, conferences, truck stops, use of right of way for telecommunications, Park & Ride fees and permit fees for oversize vehicles. IMG does not discuss de-politicized toll rate setting, one of the most powerful arguments for privatization.

IMG does acknowledge that privatization will reduce certain costs: maintenance by 20%, rehab and renewal by 20%, toll collection by 25%, overhead by 70%, policing by nothing for an overall $90m vs $115m under FDOT, a 22% saving. They say that privatized services “have routinely achieved cost savings of 15 to 40% in the operations and maintenance of public use infrastructure depending on the services provided and on the relative efficiency of the entity prior to privatization.” (p145) However they say Florida Turnpike is already among the more efficient tollroads. But the operations savings pale into insignificance compared to the assumed tax disadvantages of an investor-owned turnpike and the return on investment requirements of the private sector.

The proposal would have all employees of the turnpike enterprise designated as senior management service/ select exempt service to enable them to be paid and promoted and fired in a businesslike fashion independent of civil service regulations applying to other FDOT employees.

Over 20 years the Enhancement strategy with the FTE would see a $6.5b capital program, but the consultants say the present Economic Test of Feasibility should be ended or revised. That test presently requires each new toll project to generate sufficient revenue after operating costs (net revenue) to service half its debt by the 5th year of operation and all the debt service costs by the 15th year (FL statutes 338.223.) The consultants want to allow less financially viable toll projects to proceed, though they do not discuss the financial risks of this. A bad oversight!

Savings recommended are:

(1) adhere to federal standards only when forced to – when federal funds are being used for a project

(2) use performance-based contracting for operations and maintenance

(3) lower debt coverage ratios to leverage a larger capital program

(4) increase the scale of projects to achieve economies of scale

(5) design-build

(6) extensive use of public-private partnerships

(7) use of consultants and contractors with most core operations outsourced

Already 68% outsourced

Present “outsourced” work which includes contractor employees and consultants working fulltime for the Turnpike is 2,556 out of 3,773 or 68%. 1,135 out of 1,836 toll collectors or 62% of the total are with contractors. Concessions employ 1,055, all but 99% of the total since there are only 8 staff supervising concessions on the Turnpike payroll. Management of construction of new facilities called “production” (162) is 83% farmed out to contractors. The planning and programming (47) is 79% farmed out and public information (12) 75%. Non-tolling operations (235) which covers maintenance, ITS, tests and inspections and traffic operations is 64% outsourced with 151 contractor staff to 84 Turnpike employees. Policing (125 positions), finance (17), facilities and equipment (111) and central office functions (65) are done mostly in-house.

IMG’s analysis says continuation of the present fragmentation of the Turnpike within FDOT will limit its potential. Current FDOT procedures for personnel, procurement and project development often “slow down activity and waste valuable resources” (p137) and the separation between toll collection and tollroad operations causes “a lack of coordination and responsiveness... to the detriment of customer service.” (p137)

Florida Turnpike toll rates at 6c/mi (4c/km) are higher than Oklahoma 4c/mi (2.4c/km) while the cheapest pike is NYS Thruway 3c/mi (2c/km) while six others are more expensive – up to 19c/mi (12c/km). On average other facilities are 11c/mi (7c/km). [Two IMG number are plainly wrong. HCTRA tolls in Houston are not 19c/mi, but about 11c/mi cash and 8c/mi with ET. GSP NJ is the cheapest at 2c/mi]

Present problems

The report says the present separation of responsibility within FDOT for tolling and other aspects of Turnpike management in the Turnpike District is dysfunctional. It fragments responsibility for customer service and results in a lack of coordination and cooperation, and limits the Turnpike’s ability to promote its facilities and services. The consultants also say there is uncertainty at middle management level about the authority of contractor bosses versus Turnpike District management. Section managers work one-on-one, contractor to Turnpike, but at middle and senior management levels there is a dual reporting system up separate contractor and District command chains.

IMG identifies a number of opportunities to improve the Turnpike’s performance: more outsourcing and privatization of functions, more ‘partnering’ with other public and private entities, improved connections to other highways and modes, design-build, more maintenance contracting, advertising, branding, truck stoops, hospitality services and ‘innovative business practices.’

It says however that the Turnpike is considered an innovative organization among tolling agencies, and that there is an opportunity for the rest of FDOT to adopt many Turnpike practices.

“Threats” facing the Turnpike are listed as (1) coordinated efforts by environmental and local groups to stop expansion projects, (2) competition from non-tolled FDOT roads, (3) efforts by other districts to limit the Turnpike, (4) resistance to tolls and toll increases, (5) diversion of Turnpike funds to non-Turnpike purposes. The consultants say that the Turnpike suffers various constraints within FDOT including: pay limits that “prevent the Turnpike from effectively competing in the marketplace for available talent,” personnel processes that are “slow and cumbersome” and “high fragmented” training and development programs. (p32)

The Enhancement Option, C is described as the only one allowing it to “fulfill its mission of expanding the Florida Intrastate Highway System.”(p185) It suggests that a number of toll facilities not operated by the Turnpike District be added to the new Turnpike Enterprise including:

• Alligator Alley (I-75) Fort Lauderdale to Naples

• Bee Line East Expressway (FL-528) east of Orlando

• Navarre Bridge which links Santa Rosa Is FL-399 to the mainland east of Pensacola

• Pinellas Bayway bridges in the Tampa area

• Sunshine Skyway Bridge at the mouth of Tampa Bay

Option C, Enhancement, seems to consist of doing more forcefully what the Turnpike is already doing: leasing right of way, getting concession revenues, adding service plazas, more advertisement sales, developments along the turnpike, fees from park & ride and oversized vehicles (p187.) Operations and maintenance could be reduced by about 10% by adopting methods used by highly outsourced agencies. E-470 is cited as an example with O&M costs 10% below FDOT’s per lane-mile. A 40% reduction in overhead costs is suggested as feasible. An overall 10% lower operating expense is posited form greater efficiency as compared to the present split District/OTA arrangements.

Toll revenues are expected to grow from the present $370m/yr to over $500m/yr by 2006 when the first toll increase is assumed to $800m/yr in 2016 when the second goes into effect and to be $970m in 2020. Net present value of operating cash flow to net present value of debt service is estimated at 1.35.

The report has a huge attractive list of improvements the Turnpike could make under its Enhancement scenario. It is not clear however that these would enhance net revenues. Some might reduce them.

Under the Enhancement scheme there would be a Florida Turnpike Enterprise director who would be answerable to the secretary of transporation and with a designated member of the Florida Transp Commission. (IMG 301 907 2900)

COMMENT: This report looks more like a dream-plan for the present management of the Turnpike than an objective assessment by an independent outfit. The additions suggested to the Turnpike system are presently operated by FDOT or by counties. The Enhancement option they advocate would enhance the flexibility and freedom of action of management within the state sector without exposing them to the discipline of the marketplace. Risk and losses are still with state taxpayers. The report is unconvincing. IMG consultants present figures showing a privatized turnpike would construct less new tollroad and employ fewer personnel than the state-owned FTE. This is put as an argument against privatization.

It is arguable whether the state scheme’s extra tollroad is warranted. It implies that projects with a lower than market rate of return on investment and greater risk will be financed by the state. There is an argument first, whether such projects deserve to be financed - which IMG does not touch. But second, state ownership is not the only way to support projects which the private sector won’t finance by itself. A model being used by the Texas Turnpike Authority, for example calls for private sector proposals that involve a bid or negotiated level of tax contributions as a supplement to private sector funding. So it would be possible to have privatization plus tax supported tollroads, a model the consultants have not examined.

IMG writes as if the only possible privatization is for all eleven tollroads in the Florida Turnpike system must be sold off to a single private sector entity and all future tollroads built by that entity. Many different arrangements could have been considered. Parts of the Turnpike system could make more sense operated as parts of the regional toll systems in Tampa, Orlando and Miami. There is no discussion either of (1) privatizing financially self-supporting parts of the road system, and (2) maintaining a Toll-Assisted Enterprise within the FDOT to handle non-self supporting roads that the legislature believes justify some tax monies in addition to toll revenues.

Further, generating more employment is presented by IMG as an unqualified virtue. It is only a virtue if it generates more value-added, but value added is less with the turnpike in state hands than in investor hands, so value-added or productivity per employee is significantly lower.

The IMG report says a privatized Turnpike would have a small staff, possibly as few as 14 - most being managers - with all the line functions contracted out. It might. But IMG is rather dogmatic. There are many different private sector models. Some investor-owned operations that apparently contract out are contracting out to partially or wholly owned subsidiary companies. Others - Autostrade, 407-ETR for example - contract out rather little and prefer salaried staff. When 407 in Toronto was privatized, they ended most of the outsourcing that the provincial OTCC had in place and put people on staff.

But saying the privatized Turnpike would have almost no staff makes privatization a good ‘scare’ story with the present employees.

IMG presents the taxes a privatized turnpike would pay as a loss to the road system, as compared to the lack of taxes paid by the state Turnpike. But the road system’s gain is someone else’s loss of revenues forgone. It is a redistribution, not an economic loss. One financier we spoke with said the IMG scenario greatly exaggerates the taxes that would be paid because it overlooks the way corporate accountants can minimize taxes with various ‘creative’ but legal devices. He said investor-owned businesses face exactly the same tax problem in wastewater, electricity generation and other utilities, but somehow they manage to compete with government owned tax-free businesses.

IMG’s methodology as used to squash privatization of the Turnpike would make the case for supermarkets, banks, airlines, indeed any business at all, to be operated in the state sector. If one assumes a given demand for the product, and state enterprise exemption from taxes, no need to make a competitive return on capital, and access to some tax support for, say, state-operated supermarkets in needy areas, then certainly a state enterprise would build more supermarkets (more lane-miles) and employ more people in supermarkets than business. But we don’t have the state operate supermarkets because of a variety of factors conveniently overlooked by IMG in this bogus study: (1) risk is more fairly borne by voluntary investors than by taxpayers (2) the competitive pressure to generate return on capital produces innovation, cost-savings and a focus on customer needs that a state operation cannot (3) a state enterprise is limited by geographical bounds whereas a supermarket can take advantage of buying power, training, recruitment and knowhow across irrelevant governmental borders and may usefully be part of a regional or national ‘chain’ of supermarkets (4) a large state supermarket would despite much exhortation to innovative businesslike methods, partnering, flatter management structure, streamlining etc remain more bureaucratic, and less responsive to the marketplace than competing investor-owned supermarkets with their stock traded in the capital markets and management subject to removal by investors (5) a large state supermarket operation would by its nature be subject to close and constant political scrutiny, legislative oversight, pressure for patronage appointments, pork-barrel favors for certain districts and special interests that would divert it from businesslike decisions and structures (6) the board of directors of the state supermarket and senior executives would change each time the governor changed and liable to be driven by next election issues.

A major argument put against privatization is that it would commit the state to a fixed course of action for at least the next 20 to 25 years “thereby reducing its flexibility to react to changing conditions and trends regarding infrastructure needs and funding opportunities.” (p181) This would be true if the private entity were unable to respond to changing conditions and trends, or if the state were unable to do any tolls outside the privatized entity. Finally it assumes the state could never buy out the private operator - which is exactly what a state government did last year in New Brunswick, Canada.

It is quite unclear form the report why private turnpike management would accept almost a third of its network operating at LOS-F, as IMG suggests, unless it was bound by rigid restrictions on its freedom to set toll rates. Heavy congestion is a clear signal that motorists are willing to pay higher tolls, and at some point the higher tolls will finance the extra capacity needed to avert that congestion. The IMG modeling makes no sense unless it assumes an idiocy on the part of Florida’s government on the scale of California’s electricity policies. When it comes to Option C, IMG suggests differential peak/off-peak pricing on congested parts of the system and to “optimize toll revenues,” (p207) a policy that it overlooks in the privatization scenario.

Right at the end in the final conclusions the report does state that an “O&M Contractor Privatization” scenario ranks with the Enhanced FDOT Turnpike Enterprise in generating the highest net revenues to support the largest capital improvement program.” (p219) Actually the numbers presented (p218) show the O&M Contractor Privatization slightly better at supporting the capital program (1.5 vs 1.47 NPV debt coverage ratios) than the recommended Enhancement Option. But the true privatization option is killed by the more conservative debt service ratios assumed for an investor operation (2.1 vs 1.5) plus the same-pot revenue assumption combined with taxes (13%), dividends (5%) and royalty payments to the state (2%)

There is a major discrepancy between the text which says privatization would yield “a modest cash infusion of about half a billion dollars” (p235) and the numbers for potential purchase price of “$2.9b to $4.5b” in the tables (p215). If the table numbers are nearer the mark, and debt to be defeased with the sale of the system is $1.7b then privatization would provide a substantial pot of cash ($1.2b to $2.8b) to support other toll, or non-toll projects, or lower the gas tax. More realistic assumptions about the ability of investors to raise toll rates, and to limit their tax payments, would completely change the conclusions.

Unfortunately this study went out of its way to avoid considering the benefits of privatization while it dwells on, and exaggerates the downside. It paints a propagandistically rosy picture of the ability of a government enterprise to be businesslike within the confines of a government department. And it overlooks the value of taking the turnpike out of political control so it can serve customers rather than special political interests. It avoids looking at a mix of privatization for the viable parts of the system and provision for toll-assisted (tolls plus taxes) for other desirable roads. Nowhere are the restrictive assumptions justified. They are simply asserted.

The whole report reads like a put-up job: “Here’s the obvious conclusion, guys. Go find some arguments and data to support it.”

BACKGROUND: Florida’s state turnpike system consists of a 512km (320mi) ‘Mainline’ that goes from the southwest of the Miami area to Wildwood northwest of Orlando and includes the Bee Line West spur to the east coast, and seven new tollways, totaling 197km (123mi) opened from 1990 onward. Two of these are in the Miami area, where four other tollways were taken from the state and transferred in 1997 to the jurisdiction of the Maimi-Dade Expressway Authority (MDX), and five in the central Florida area. This includes the latest and largest, the 67km (42mi) Suncoast Parkway north of Tampa which is actually two-thirds open. The 10km (6mi) Seminole Exwy 2 is under construction in the northeast of the Orlando area, and the state is developing an 18km (11mi) southern segment of the Western Expressway (FL-429) in the southwest of the Orlando area. There is no turnpike activity in northern Florida. By 2003 it is expected to be a total of 718km (449mi). State turnpike constitutes 9% of the miles of the state’s so-called Intrastate Highway System (FIHS) which despite its name encompasses all the Interstate highways as well as other motorway standard roads. It carries 12% of the FIHS traffic and has been about a fifth of the growth in lane-miles, and virtually all the new right-of-way highway of that system. It generates about $330m in tolls on about 900k toll transactions per day and $10m in concession revenues. The toll system is a trip-based ticket system in the rural stretch between the east coast and the approaches to Orlando but otherwise is a point toll system involving barrier and ramp plazas.

Tolls average 4c/km (6c/mi) for cars and 11c/km (18c/mi) for trucks. The legislature in Florida has generally allowed the Turnpike to price its tolls at realistic levels and encouraged a considerable degree of contracting out of operations which has kept costs low. As a result the turnpike system has among the highest bond ratings in the country (Fitch A+, S&P AA-, Moody Aa3.)

Florida’s turnpike was built by a Florida State Turnpike Authority (FSTA), independent of the DOT. Founded in 1953 it built the original Mainline from the northern part of the Miami area 427km (265mi) to its present terminus at I-75 northwest of Orlando. The first 176km (110mi) section Miami-Fort Pierce opened 1957 and the full mainline to Wildwood opened 1964. In 1969 the FSTA was disbanded by the legislature as part of a Reorganization of State Government Act and its functions taken over by the FDOT. Toll collection and finances were centralized but new construction and maintenance was decentralized to be handled by the regional districts in which the work was located. In 1988 legislation provided for an Office of Florida’s Turnpike (OFT) within FDOT to concentrate Turnpike non-toll operations. OFT was redesignated Florida Turnpike District in 1994 and to this day it conducts non-toll functions, while OTO manages toll functions. The Turnpike District is headed by Jim Ely (407 355 5733), and OTO by Deborah Stemle (850 488 5687)

There were toll increases in 1979 and 1989. The 1989 increases were linked to a new expansion plan, authority for bond issues, and the economic feasibility test designed to limit non-viable projects.

1974 saw 88km (55mi) of extra tollroad opened: an Orlando spur, the Bee Line West and a Miami area western peripheral, the Homestead Extension of Florida’s Turnpike (HEFT.) In the 1990s openings included a 37km (23mi) northwest peripheral in the Miami area, the Sawgrass Expressway (1990), a 19km (12mi) Seminole One Expressway (FL-417) in northwest Orlando (1994), the 24km (15mi) Veteran Exwy norht of Tampa (1994), a 10km (6mi) Southern Connector in Orlando connecting I-4 to the OOCEA’s FL-417 Greeneway (1996), and a looproad around Lakeland off I-4, the FL-570 Polk Parkway (1998/9.) At the end of 1999 the Turnpike was 645 centerline-km (401mi) about 10% of the state’s motorway standard and major arterial roads (FIHS) with 2,722 lane-km (1701 lane-mi) or 11% of the total capacity. Vehicle-miles traveled were 12m daily (19m veh-km) 12% of that carried on the whole FIHS system. The Mainline does 75% of the 910k/day transactions and 81% of the revenue. 96% of traffic is automobiles, 86% of revenue.