That would mean a total of a little more than 1.2 million jobs were created or saved in 2009, versus the Council's forecast of 1.5 million to 2 million jobs, though comparisons are difficult.Comparisons are made even more difficult by the deadweight loss associated with any subsidy. In the picture below, the deadweight loss in the job market is given by the red shaded area (click here to enlarge).
It is important to specify that the government has taken money from people to "create" jobs, redistributed it, and claimed victory, likely at the cost of future jobs. If the government had not decided to push forward with a stimulus, the money they no longer needed could be re-invested by consumers in more productive activities--hopefully, in spending that does not generate deadweight loss. Natural investments, which spur industrial growth, may create many more jobs in the future than the government has burdened tax-payers with now.
It is also important to note, as Kevin Murphy has, that the effect of government spending depends largely on how efficient the government is at employing the resources it has. If we believe that the government is largely inefficient (so for every tax dollar it raises, it entirely wastes fifty cents), this makes it more likely that the net value of the stimulus is negative. If we notice that the stimulus plan is not devoted to making use of "idle" resources, but rather to transferring funds from productive taxpayers to less productive citizens, this also makes it more likely that the net value of the stimulus is negative. Finally, if we believe that even "idle" resources, such as the unemployed, are able to become productive with relative ease, such as by taking a college course or moving, the case that the stimulus has net negative value becomes quite compelling.