Gov. Rendell is an optimist. Not a math major.
He says he can sell toll-taking rights on the Pennsylvania Turnpike for a fat up-front payment, pay off its debt, invest the remaining $10 billion and change with the state pension fund, set aside a bit for inflation, and pocket an average 12 percent each year, or $1.1 billion, to fix rundown roads and bridges.
Why stop at 12 percent? Why not take it all to Harrah's in Chester, and go for a jackpot?
"It's not a guess," insists Roy Kienitz, the bow-tied deputy chief of staff Rendell used to defend his math.
According to the governor, the State Employees' Retirement System has posted average returns of 12 percent a year, for the last 20 years. And if it can do that for retired senators, judges and social workers, Rendell figures it can do that for asphalt contractors and steel fabricators hired to fix roads and bridges.
There's a lot riding on that 12 percent guesstimate. For every percentage point missed, the roads are short nearly $100 million. If the yield stays below 10 percent, the state misses its yearly target and has to find other funds or scrimp on repairs.
In real life, SERS pays pensions with help from hundreds of millions in taxes and state worker "contributions" each year. SERS's reported investment returns depend on estimates of the value of its unusually large pool of hedge funds, real estate, venture capital, and other private assets, which it might not be able to sell for years.
Tying future highway spending to those returns is a little less scientific than betting your house will go up in value, from now until your kids are dead, at the same rate it rose from the end of the Reagan presidency until the real estate market keeled over last year. (SERS spokesman Bob Gentzel adds that returns weren't quite 12 percent - "for the record, 11.5."
Looking forward, SERS's own profit target is just 8.5 percent, which would cut a $320 million yearly hole in Rendell's budget. SERS might not meet that goal either, chairman Nick Maiale reminded his ex-colleagues in the General Assembly earlier this year.
That didn't move Kienitz. "Pension fund managers, their goal is to underpromise and overdeliver," he said.
Let's ask some outside experts.
"Even an 8.5 percent return assumption is generous," said Ted Aronson, who manages almost as much money as SERS does through his Center City investment firm, Aronson + Johnson + Ortiz L.P. "It's one thing to be optimistic about the markets; it's another to believe in the tooth fairy."
"The Securities and Exchange Commission challenges assumptions over 8 percent" for corporate funds, said professor Olivia Mitchell, a pension scholar at the Wharton School. "Even 8 percent is fairly aggressive. Some people in this building think returns on assets will be a lot lower in the future than they've been in the past."
Kienitz doesn't buy any of that. "I understand the industry uses a lower number," he said. "Do you trust some guy picking a number out of the air? Or do you trust the data?"
He told me the state's consultant, Morgan Stanley, backed the governor's math. The guy who crunched those numbers didn't want to talk about it. But his firm's 2007 report, helpfully linked to Rendell's Web site, includes a chart estimating investment yields at 7 percent to 9 percent. Rendell's 12 percent estimate came on further reflection, Kienitz said.
Back on Wall Street, Citigroup Inc.'s Felicity Gates, a leader of the partnership that made the winning bid of $12.8 billion in last week's turnpike auction, said a bigger concern is the General Assembly's lack of enthusiasm, which makes this deal "really unusual."
"When we did the Chicago Skyway, the legislative support was there," she said. "With the Indiana Toll Road, the legislation hadn't yet been passed, but it had support. In Pennsylvania, the legislation doesn't have the support it needs yet."
What's her theory? "It's partly the way it works here, the way politics is conducted," she said.
Not that Gates is complaining. This would be, by far, her group's biggest highway. It's worth some risk.
"An open and frank debate is a good thing," Gates concluded. "I'd rather see it happen now than when it's nearly done.