Gasoline Consumption Forecasts Keep Dropping

Last week, our “Peak Gas” report highlighted the fact that Washington state officials now predict a long-term decline in total gas consumption. But already, the state has put out a brand new transportation revenue forecast. The big news: for at least the fifth consecutive quarter, the state’s long-term gas consumption forecast has dipped.

The red line in the chart represents the most recent projection, stacked up against the previous eight forecasts. The latest numbers show a nine percent decline in annual statewide gas consumption through 2027.  That’s down significantly from the June projection, which was lower than the quarter before, which was lower than the quarter before that…which was dramatically lower than any prior projection.

The question right now in my mind is: when will the declines stop? After all, some big-name energy forecasters are predicting much steeper declines than nine percent for the US as a whole. Deutsche Bank recently said that US consumption would decline by about one-third by 2030, while the ever-optimistic Cambridge Energy Research Associates has pegged the long-term decline at “only” 20 percent over the same period.

So it’s possible that, if anything, OFM is still being optimistic about future gas tax revenues.

If you look past the “holy cow” message of the red line, perhaps you’ll wonder what happened in November 2010 to make the forecasts change so dramatically.  The answer is pretty simple: the OFM radically revamped the model they use to forecast gasoline consumption. They had to: the old one wasn’t working at all.

Here’s a chart that illustrates the point. It shows actual gas consumption trends in pink (I take the trendline from federal data, and adjust it a bit to match up to the state figures) against the six most recent official forecasts. It’s pretty obvious that, given a decade of stagnation in gas consumption, the OFM forecasts prior to November 2010 (the ones in orange and yellow) were ridiculous. As were all of the forecasts they’d been making for years. The OFM itself admitted the error, noting that the real-world trends were more than two standard deviations below their previous  predictions.

So if you look beyond the grim implications of the red line for highway finance, these two graphs demonstrate a powerful lesson: official forecasts are just guesses, and not very good ones at that.

Yet to a profound extent, these guesses—about traffic volumes, fuel prices, economic growth, tax revenues, the policy environment, and so forth—shape the conventional wisdom and public debate about what highway investments make sense over the long haul. But in the last decade or so—ever since oil prices started rising—the official projections have proven simply laughable. Nothing we thought would happen has come to pass: rather than growing inexorably, both gas consumption and VMT have flat-lined for a decade.

To me, the lesson here isn’t simply that we need better forecasts. Instead, it’s that if we made our transportation investments differently, we could free ourselves from the pitfalls of forecasting altogether. Rather than making big bets on highways that we may not need, and probably can’t afford, maybe it’s time to consider a different way of investing in transportation: small, flexible, can’t-miss investments in refurbishing the streets, bridges, and roads we already have. Those kinds of incremental investments could make for better neighborhoods, and set ourselves up for all kinds of transportation futures—without forcing us to bet the farm on the riskiest sorts of megaprojects that can only work if our forecasts are perfect.