Michael Ennis may have batty ideas about parking policy, but he’s got a nice post today on a subject I’ve been meaning to write about: Washington state transportation revenue is in dire straits. New models show that revenue forecasts have consistently missed the mark.

To illustrate the problem, check out this analysis of gasoline consumption, which is (obviously) a key driver of gas tax receipts:

revenue actuals

What it shows—if you look at the difference between the solid blue line and the solid pink line — is that actual gasoline consumption under-performed the mean forecast by more than two standard deviations over the last 17 years. Yikes.

  • Factor in that change and some other revised modeling assumptions, and the new November 2010 revenue forecast slices $1.7 billion off the revenue forecast that was produced just two months earlier. (Granted, that reduction is over a 16 year horizon, but that’s an awfully huge reduction.) What’s worse, I’d argue, is that even the new sober forecast still appears to be premised on some dubious assumptions, like this one.

    Can anyone tell me what’s odd about this chart showing the current projections for gasoline consumption? If you guessed, “it forecasts immediate growth in consumption starting next year despite a) recent history; and b) economic conditions” then you are correct.

    I’d say that’s an unduly rosy scenario for gas tax revenue.

    All which has real-world implications. Call me a conservative if you like, but I don’t think we should embark on expensive megaprojects that we can’t pay for. The state highway system is riddled with pressing needs for basic maintenance, preservation, and safety. It seems only prudent that we spend our limited transportation dollars on those necessities, rather than on huge new highway segments—like 520, the CRC, and the deep-bore tunnel—that are, in aggregate, billions of dollars short of funding, and premised on Pollyana assumptions about cost control.