Stay up to Date
Our weekly insiders' guide to political and policy news (Saturdays)
The San Diego Association of Governments is on track to collect billions of dollars less than officials said it would from a sales tax hike voters approved 12 years ago, throwing into question whether it will have the money to pay for many of the projects that measure promised.
The causes of the shortfall, unreported until now and only disclosed in a complex budget analysis document, could also jeopardize many of the promises in another sales tax hike SANDAG is pushing on this November’s ballot, Measure A.
SANDAG leaders obscured the size of the shortfall in the one place they disclosed it. They have not told the public of risks that they won’t get the $18 billion they say Measure A would generate over 40 years, meaning some of the promised projects in the ballot measure may not get funding.
“It’s very significant, because voters are being misled, possibly deliberately or certainly with negligence,” said Peter Kiernan, an attorney with Schiff Harden who specializes in public finance and infrastructure and who worked as special counsel on those issues for the city and state of New York.
“If your revenue modeling is proven wrong, you have an obligation to change it. And if they’re not willing to admit that it’s been proven wrong, then they’re not looking at the facts,” he said.
Voters in 2004 approved an extension of TransNet, a half-cent sales tax to fund freeway widenings, light rail expansion, habitat preservation, bike lanes and road improvements across the county. The tax went into effect four years later. So far, SANDAG has collected 25 percent less revenue under TransNet than it promised voters on the ballot.
If sales tax revenue grows at the rate it has grown since the end of the recession, TransNet would end up bringing in around $9 billion for transportation projects – or $5 billion less than voters were told.
SANDAG officials acknowledge they have not made any substantial changes to the agency’s forecasting methodology since then — including in the forecast used to project $18 billion in new revenue from Measure A, another half-cent sales tax increase facing voters on this November’s ballot.
Campaigns for and against Measure A are focusing on how the $18 billion it’s expected to raise will be spent – whether enough of it goes toward transit, for instance, or if North County receives its fair share. It turns out there are serious questions over whether the $18 billion they’re fighting over will materialize at all.
A SANDAG official also confirmed that money from Measure A could be used to backfill any shortfall from the previous tax – though he said he’s confident that won’t be necessary. But there’s a chance that voters weighing in on Measure A could be approving a tax to pay for projects that were already supposed to have funding, not the new projects that are the subject of the current campaign.
When SANDAG started collecting the extension of TransNet tax in 2008, the economy had cratered from the Great Recession and revenues were already way below expectations.
Tax returns improved after the recession ended – though not enough to approach initial estimates. For the last two years, tax revenue growth has actually declined, meaning the gap between SANDAG’s initial revenue expectations and reality has widened recently.
To date, SANDAG is $500 million short of the $1.9 billion it expected to collect by now. If revenues grow at the average rate they have since the recession, they’d collect $9 billion over 40 years, not the $14 billion promised to voters. That would mean $5 billion less to spend on new regional transportation projects.
SANDAG’s chief economist Ray Major said the revenue shortfall is merely a function of the recession.
“While revenue receipts tend to fluctuate in the short term, in the long term, peaks and valleys tend to even each other out,” Major said in a written response. “Currently, we are just a decade into the TransNet extension so it’s premature to conclude that by the end of the measure, we would be short of funds to complete our projects.”
Marney Cox, the agency’s special projects director and former chief economist, said SANDAG doesn’t know how much money it now expects to collect through TransNet. He said it’s immaterial whether it brings in the projected $14 billion, the $9 billion it’s on pace for or some number in between.
“At this point in time, it’ll be sufficient to pay for all the projects, so you can put whatever number you want on it,” Cox said.
SANDAG claims the primary reason for its lagging revenue is the Great Recession.
Other experts aren’t so sure.
For one, SANDAG’s sales tax revenues have improved since the recession ended, but they’re still nowhere near the agency’s initial expectations.
Throughout the country, state and local governments are noticing slow sales tax growth, and project the sluggishness will continue. A survey of state forecasts from the Rockefeller Institute of Government, a public policy research group within the State University of New York, said states anticipate a slowdown in sales tax growth in 2016 and 2017.
That’s why outside experts aren’t as comfortable as SANDAG that the low sales tax returns will take care of themselves over the next 30 years.
“It is multiple items that are leading to a really weak recovery in sales taxes,” said Lucy Dadayan, a senior policy analyst at the Rockefeller Institute. “It’s observed all across the country. (SANDAG’s) claim about this being from the recession is partially true, but it’s also a changing nature of our reality.”
Dadayan said sales taxes since the recession haven’t rebounded nearly as fast as income taxes. She said it is possible consumer behavior changed due to the recession, but states are also losing out due to low gas prices and an increase in online shopping. That’s meant many states and local governments no longer get a cut of their residents’ purchases.
SANDAG’s Cox said San Diego has seen sales tax collections fall for two region-specific reasons, too. The collapse of the Peso has cut down cross-border spending, he said. And as the housing market pushed some residents to relocate to southern Riverside County, he said, some of their daily shopping now happens outside the county as well.
SANDAG’s forecasting model didn’t account for any of those things in 2002. And they haven’t been incorporated into the model that was used to project that Measure A would generate $18 billion.
Cox said that’s a good thing.
“You have to be patient about trends,” he said. “We end up being a little patient before we recommend permanent changes to the way we do things.”
Bill Glasgall, director of the state and local program at the Volcker Alliance, a foundation focused on improving the performance of government at all levels, said he’s skeptical that the shortfall should be primarily attributed to the recession.
“In general, the recession is fading in our rearview mirror,” he said. “But sales tax growth in general has been positive but weak. It’s basically flat. The driver for that weak sales tax growth is internet sales and low gas prices, and nobody projected this.”
The TransNet package voters approved included something meant to bring comfort to residents parting with more tax money – an independent watchdog group that would look after the funds and ensure they were being spent appropriately.
SANDAG disclosed the revenue shortfall in a document submitted to the Independent Taxpayer Oversight Committee.
But in a number of ways, the document obscures the straightforward fact that SANDAG isn’t collecting as much revenue as it told voters it would.
If you were to compare the revenue SANDAG predicted would come in when Transnet was approved with the revenue it’s actually collected, the shortfall would be clear. SANDAG didn’t do this. Instead, it used a different set of predictions that it adjusts each year. When the revenue actually collected is compared against those yearly estimates, the shortfalls appear much smaller.
But in a paragraph just below that table, SANDAG staff describes how the tax isn’t bringing in as much money as they said it would on the ballot.
In other words, the paragraph describes something entirely different than the table that precedes it.
SANDAG diminished the size of its revenue shortfall in another major way too.
It didn’t compare revenue collections to the $14 billion voters were promised. It adjusted the forecast for inflation to the first year it started collecting the tax, 2009, which reduced the size of the shortfall to about 12 percent, instead of 25 percent. Nowhere in the oversight document did SANDAG compare its actual revenue to the original $14 billion projection.
Cox defended that way of accounting for the agency’s revenue shortfall.
“You have to know what the revenues would be in 2008,” he said. “When we got there, we had to make the adjustment.”
Richard Carson, an economist at University of California at San Diego, said it’s perfectly reasonable to measure the shortfall from the original $14 billion that appeared on the ballot, because that’s what voters were promised. He said SANDAG shouldn’t be faulted for over-estimating how much revenue it would bring in, but the fact is it has now collected 25 percent less revenue than it told voters.
“The common sense way to do it is, ‘What did the voters see?’ and that’s the way you’ve done it,” he said. “The fact that they were off, well everybody who was forecasting these things was off. And they’re trying to say, ‘Well by the time we started collecting money, we had adjusted and we weren’t so surprised.’ But it’s still about what people would have been making their decisions on, and what they’re saying doesn’t change any of that.”
Kiernan said you can’t blame SANDAG for getting the forecast wrong, because “the land shifted under their feet,” but he said the agency deserves plenty of scrutiny for refusing to openly acknowledge the severity of the shortfall.
“Given the facts, they should in their periodic disclosures, their continuing disclosures and official public statements, they should be telling this story,” he said. “They have to come clean on this. It’s very dramatic.”
The revenue shortfall carries clear consequences.
Multiple experts said SANDAG would have to build fewer projects than promised, or find money from another source to make up the difference.
Plus, they said, the shortfall itself threatens to erode public trust.
“They promised people a set of improvements,” Carson said. “Now they either have to cut back on the improvements because they’ve taken in less revenue, or they have to pay for the revenue shortfall from another source.”
“To continue building the planned transportation infrastructure, they’ll have to take money from elsewhere,” he said. “That’s if they continue projects as planned, as opposed to just abandon them.”
“It really isn’t complicated,” said Robert Puentes, president of the ENO Center for Transportation, a think tank focused on transportation policy that is currently monitoring regional transportation sales taxes across the country. “They have to roll back their ambitions, and they have to manage public trust. Eighty percent of these ballot measures passed because people are willing to invest in infrastructure when it’s transparent. If those promises aren’t met, it’s going to whittle away at public confidence.”
“We ask a lot out of individuals to make these decisions, and it’s harder to ask them to do that without giving them all the information,” said Sarah Swanbeck, executive director at UC Berkeley’s Center on Governing and Investing in Our Future, who is currently working on a transparency in budgeting project.
One additional source of money that could make sure projects promised in 2004 are still finished is the new revenue from Measure A, if it’s approved.
Cox acknowledged that could happen, though he said he doesn’t think it would – for two main reasons.
One is that SANDAG received more in state and federal funding for its projects than it assumed it would under TransNet. That’s meant building more projects than officials might have expected so far, given the size of the shortfall.
But counting on that going forward is risky, multiple experts agreed.
For one, it counts on political compromises at the federal level, even while the two parties are as far apart on policy as they’ve ever been.
“The federal-state-local relationship is just being completely upended,” Puentes said. “A major reason a lot of places are raising money at the ballot locally is so they don’t have to deal with the federal government.”
For another, it means competing with other regions for a finite amount of money.
“You can say, ‘We’ll try to do this’ and have a reasonable amount of certainty, but it’s still discretionary money, so there’s no way of knowing you’ll get it,” Kiernan said. “The secretary of transportation can either be fired, or become the secretary of the treasury. And even if you get the money you want, it may not be the year that you want it.”
Local money that you control is better than federal or state money that you don’t. In fact, SANDAG says as much in the ballot language for Measure A, which it describes as “a 40-year, half-cent sales tax ($308 million annually) that Sacramento cannot take away.”
The other reason Cox is adamant that SANDAG will still meet its promises to voters is that the agency aggressively took advantage of lower construction costs in the early years of the measure, hoping to get more done with every dollar it spent.
“We were out there getting as many projects under contract as possible, because prices were exceedingly low,” he said. “The fall in prices is the most important thing, because our purchasing power was elevated.”
Indeed, the state’s construction cost index fell nearly a quarter from 2007 to 2009 and stayed low through 2012. But construction costs have come roaring back and are now at about their 2007 peak.
In the end, Cox’s argument for why voters should expect SANDAG to complete all the projects it promised in 2004, without dipping into the new money potentially secured by Measure A boils down to: Trust us.
The 2004 vote was an extension of TransNet, which was first approved in 1987. SANDAG built all but four projects in that measure, and those four projects were put into a lockbox and prioritized in the extension. One of those is the trolley extension from Old Town to University Town Center, which broke ground Saturday, just shy of 30 years after it was initially approved.
“The true measure of this is, when you ask taxpayers to extend the program, and two-thirds of them say yes, and so they must have thought we did pretty good,” Cox said.
But Measure A is different. Voters have to approve a 40-year tax with 40 years of new spending, without the benefit of seeing how the TransNet extension performed in the end.
Cox thinks SANDAG will be able to make up for the missing TransNet revenue by 2048 and build all the projects it promised.
He thinks voters should take a leap of faith that he’s right.
There’s reason to believe Measure A itself won’t bring in the $18 billion voters are being promised, regardless of what happens with TransNet.
The forecast the agency used to produce that total – which didn’t undergo any major changes since 2002, despite SANDAG’s revenue shortfall – has an aggressive view of how much San Diegans are going to spend in the coming decades.
The agency’s forecast says the amount that the average San Diegan spends per year, adjusted for inflation, is going to rise and keep rising for the foreseeable future, surpassing the highest point it ever reached since 1970 and nearly doubling the county’s historic average.
The typical San Diegan has spent about $15,000 a year on stuff that’s subject to SANDAG’s sales tax. That number surpassed $17,000 in 1978 and again in 2006, and has been as low as $13,000.
SANDAG expects to achieve a new all-time high by 2025, and then keep climbing. By 2039 it expects that number to reach $21,000, or about 30 percent above the historic average.
If the county doesn’t reach that unprecedented spending level, then SANDAG won’t collect the $18 billion its promising voters, and won’t be able to build all of the new projects that would pay for.
“Wow. Now that’s extraordinary,” Puentes, the president of the transportation-focused think tank, said when he saw the chart of the agency’s expectations for local sales.
“Typically you can’t expect much more than the historical average,” Dadayan, the sales tax-focused researcher, said. “And you should be very conservative, usually less than the historical average. It’s better to collect more money than to underperform.”
Cox stood by the model and its expectations. The growth rate it expects isn’t as steep as the one the county saw from 1970 to 1978, he said, and is comparable to the one it saw in the mid-‘90s.
If voters approve Measure A and the agency ends up being wrong, SANDAG would again be collecting less tax money and building fewer projects than it promised voters.