Analysts have long recognized the tendency of companies who are forced to report bad news to make the news worse than they have to, piling every single thing that might goi wrong into one hell of a charge-off. The logic of this is simple: if your stock is going to take a hit, make it one gigantic hit, so that you can later "surprise" everyone when aliens from the Planet Zork do not actually land, vaporize 2/3rds of your customers, and keep the rest too busy dodging laser rays to focus on purchasing your product.
Looking through Obama's budget, I am reminded of those massive one-time-write-off festivals. Only the Obama administration has gone one better: he has actually gotten everyone to congratulate him for his breathtaking honesty.
Take the Iraq war. We were not, under any administration, going to spend as much in 2015 as we did in 2005. But by treating that spending as an ongoing cost, Obama now gets to take as much credit for reducing it as he would for closing permanent air bases in Germany, or trimming Social Security. Reducing the cost of "overseas contingency operations" acounts for $1.5 trillion of Obama's much vaunted $2 trillion in savings. Likewise the AMT fix--with high-end incomes falling, deflation in the air, and homeownership rates declining, AMT collections are going to decline even without a fix; this lets them recognize the entire decline at a time when the numbers are so large that taxpayers are too dazed to notice the fall.
The Obama administration is hardly the first to calculate the numbers to allow them to deliver upside surprises; during the Bush administration, the forecasts issued by the White House's Office and Management and Budget started to diverge from those of the Congressional Budget Office in an unexpected direction: they became markedly more pessimistic. Few people think it was an accident that this allowed the Bush administration to deliver a steady stream of "Surprise! The budget deficit is falling even faster than expected!" announcements.
Obama needs those big bath numbers on the Iraq side, because it seems unlikely that a lot of the things he's counting on to bailout his budget are going to materialize. Health care savings are often promised by American politicians, but so far never delivered. The cap and trade revenues which are supposed to deliver $625 billion over the next 10 years are going to be politically controversial, and also, highly dependent on energy demand--if there are too many permits, they won't yield much revenue.
Indeed, though Obama is getting a lot of credit for projecting out ten years, I'm not sure I see the point from an economic, rather than a political, perspective. These models consist of the modellers assuming that most things will continue to be largely as they are, unless the economy is in a recession, in which case everything will return to 3% trend growth 18 months after the recession started. They don't track reality all that much better than you would by throwing darts at a board with a bounded range of GDP growth around a mean of 3%. Given that we are in most unual times, they are particularly suspect now.