The study states that their goal is to study the question: "What is the gross economic impact of Social Security benefits on the U.S. economy?" But this is an irrelevant question to ask, as the "net economic impact" is the relevant issue.
Social Security benefit payments in 2012 supported:
About $1.4 trillion in economic output (goods and services)
Just over 9.2 million jobs
About $774 billion in value added (gross domestic product)
More than $370 billion in salaries, wages, and other compensation
Tax revenues for local, state, and federal governments exceeding $222 billion, including $78.9 billion in local and state taxes and $143.3 billion in federal taxes
Every state—big and small—feels the effects of Social Security benefits being spent
within its borders. Not surprisingly California, with the largest economy of the 50 states,
showed the biggest impact. In California alone, Social Security benefits supported
888,000 jobs, $147.4 billion in output, and $8.7 billion in state and local tax revenues.
When I discuss examples of poorly constructed (and/or biased) economic impact studies in the future, I will likely use this as an example. This study implies that these benefits from social security arrived out of thin air. Of course, we all know this isn't true.
Each individual who works pays over 12% of his/her income (between the worker share and the share the employer pays) in social security taxes (until the income threshold is hit). Therefore, this study could also say the same thing but in reverse. You could come up with a report that said that the social security taxes cost America ...
* About $1.4 trillion in economic output
* Over 9.2 million jobs
Overall, the net effect is approximately zero, although its probably negative due to the negative distortions from government interference.
This type of study is a good reason why I tell reporters they should always be initially suspicious of economic impact study estimates. Many are done well, but too many are simply propaganda pieces (with biased estimates) designed to mislead the public.