So why have the Administration's forecasts erred in one direction only? Partly it is due to the political necessity of arguing that your policies will make things better, but a large part of it is that the Administration's economists and, by extension, the President simply have little idea how the economy works. As Keune puts it
The White House’s predictions always assume that all of its budget proposals are enacted. Then, the administration employs cheery assumptions about the effect these policies will have.
For instance, the administration’s predictions about the stimulus plan assumed a $1.55 “multiplier,” meaning that for every dollar injected into the economy, $1.55 in economic output would result. Accordingly, $787 billion of stimulus would lead to a $1.2 trillion increase in GDP.
But there’s lots of debate over the precise amount of the government-spending multiplier. Plenty of economists think that it’s actually less than one. They contend that some taxpayers save more in the short term in response to government stimulus, apprehensive that higher government spending will require tax increases in the future. In fact, the FY 2013 budget acknowledges the recent spike in the savings rate, though the administration never entertains the idea that this might be the effect of its own policies. In any case, the Recovery Act didn’t lead to the growth figures its proponents predicted. Some research suggests it actually caused a net job loss.