Well, having worked in travel demand forecasting--a combination of bad statistical models, bad assumptions about travel, and the inherent problems with extrapol

Well, having worked in travel demand forecasting--a combination of bad statistical models, bad assumptions about travel, and the inherent problems with extrapolating to produce forecasts. Travel modeling is very useful, but long term forecasts can't really account for fundamental market shifts. In my region, Central Puget Sound, long term forecasts have been thrown some pretty big monkey wrenches since people started trying to use numbers with cars. Women entering the workforce, later Microsoft becoming a thing, etc. Oddly, some historic forecasts actually turned out reasonably well... probably by accident.