SANDAG staffers tried to be forthcoming about the agency’s forecasting failure in an early draft of an op-ed eventually published under SANDAG board chair Ron Roberts’ name, according to emails obtained by Voice of San Diego. The final product obscured the extent of the agency’s error, and introduced an inaccurate claim.
SANDAG staffers tried to be more forthcoming about a fatally flawed forecast they had used to project tax revenues, but in the end the agency settled on a message that misled residents and included a major error.
The chair of the agency, who authored the misleading explanation, now says he would not have written it like he did, based on what he now knows.
In January, SANDAG board chair Ron Roberts reassured voters in an op-ed that the agency could still finish the transportation projects promised in a 2004 tax increase
We obtained public records that now show that, a week earlier, SANDAG’s executive director Gary Gallegos had sent an earlier draft of the op-ed to Roberts for review.
That draft dealt more directly with the agency’s flawed forecast than the one Roberts eventually published.
Roberts’ submission also included a factually incorrect claim – that the agency did not know whether the error had any impact on revenue expectations for the agency’s existing sales tax, TransNet, which was approved by voters in 2004. That claim was not included in the initial draft sent to Roberts for approval.
A spokesman for Roberts said the first draft was too bureaucratic and technical and he wanted it to make it easier to understand.
The piece reflected Roberts’ understanding of the situation at the time it was published, said Tim McClain, Roberts’ spokesman.
“Had Ron written that op-ed today, with the hindsight of today, it would be different,” McClain said. “That’s how he understood it at the time.”
Public records provided to Voice of San Diego don’t include any additional insight into edits the op-ed underwent before it was submitted for publication, or who requested the changes. They only include the email from Gallegos to Roberts, with the original op-ed attached.
After Roberts intervened, the new piece introduced a new argument altogether: That it is unclear whether or not SANDAG overestimated how much revenue it could collect from TransNet, a tax-and-spend plan approved by voters in 2004.
Staff had not included that claim. Instead, staff members included this line:
“After an intense effort, staff identified the root cause errors in (the forecast) in November, and they will be corrected as part of the update to the model,” read the original op-ed draft.
Roberts’ final op-ed was organized into a series of rhetorical questions. One of them was, “Has SANDAG overestimated the revenue it is likely to receive from the existing TransNet program?”
“Maybe, but we won’t know for sure until 2048,” the published op-ed reads.
Roberts chose not to acknowledge the “root cause error.” That line is completely omitted from the final op-ed. It went with this instead:
“After a month of intensive investigation, it was discovered that the growth rates of economic output for certain employment sectors used in the model were aggregated incorrectly from the source data causing the overestimations of taxable retail sales,” read the final op-ed submitted for publication.
The final op-ed also includes a couple lines that have since been revealed as categorically false.
In February, a month after the op-ed was published, Voice of San Diego revealed through internal SANDAG emails that staffers discovered the overstated sales figures and panicked over the implications, with the agency’s chief economist writing “omg” and “wtf” when he learned of the issue. Staffers then prepared a lengthy presentation for agency executives explaining the problem they found.
Specific slides in that presentation clearly spelled out that the error meant revenue expectations for TransNet and Measure A were too high.
Yet, the final op-ed Roberts submitted said the agency didn’t know whether or not TransNet or Measure A’s revenue forecasts were affected.
“Before the election, technical staff had not discovered how, or if, the model’s over-estimates could have affected the Measure A revenue forecast,” the final draft reads.
“Nor was it clear what, if any, impact that might have on the TransNet revenue estimates,” it reads earlier on.
Neither of those claims is true, and neither of those claims appears in the original draft.
The original draft says only that the specific source of the problem was discovered after Measure A had already failed.
“While, in general, staff was aware that (the forecast’s) taxable retail sales forecasts were aggressive at the time the $18 billion estimate was generated, they were considered reasonable,” it reads. “The specific issues were not identified until after voters went to the polls on Measure A.”
That last part is closer to true.
SANDAG staffers did not know what was causing the errant forecast in the fall of 2015, when they first discovered the problem. They did, however, know right away how the problem related to revenue estimates for both TransNet and Measure A. SANDAG staff quantified the impact the error had on TransNet revenues and put it into a slide included in a presentation to agency executives.
The drafts of the op-ed have one thing in common, though. Both of them discuss the fact that the $17 billion funding shortfall facing TransNet is also due to the $8 billion increase in all the unfinished projects included in the program, and not just the downward revenue revisions.
Neither of them, however, mentioned that the $8 billion cost increase was known more than a year before SANDAG incorporated it into its “plan of finance” for TransNet. Failing to do so meant SANDAG spent the entire year it was campaigning for Measure A acting like TransNet cost $8 billion less than its own official estimates suggested.