The Economic Impacts of the Advanced Encryption Standard (2018) [pdf]

Government at its best - they are claiming benefits from organising everyone to use one easily available standard for encryption.

The claimed value seems very plausible. The alternative was presumably a world more like the one where PGP's Zimmermann was being investigated for providing strong encryption. There is a real opposition in the halls of government to strong encryption and overcoming that is in itself praiseworthy.

For the discerning modern information consumer turned off by the 149 page format, here's a similarly dry reproduction in bitesize form:

~~~

Net Present Value of Benefits 1996 (NPV96):

Most Conservative Outcome: $844,500,000

Ranges for Other Infratechnology Impact Assessments: $3,500,000 - $773,000,000

~~~

Net Present Value of Benefits 2017 (NPV17):

Extrapolated Outcome: $8,772,000,000

Economy Wide Extrapolated Outcome: $250,473,200,000

~~~

Benefit to Cost Ratio (B/C):

Most Conservative Outcome: 29/1

Ranges for Other Infratechnology Impact Assessments: 3‐113/1

~~~

Table ES-1. Economic Impact of the AES program, 1996-2017

~~~

Table ES-1 summarizes the quantitative findings of this economic impact analysis in terms of Net Present Value (NPV), Internal Rate of Return (IRR), Alternative Internal Rate of Return (AIRR), and Benefit-to-Cost ratio (B/C). Looking only at the column of Table ES-1 entitled “Most Conservative Outcome,” based on the information discovered in the survey conducted for this project, in 1996 NIST AES program managers would have valued the AES program at $844,500,000, net of costs incurred. By comparison, of 12 other NIST infra-technology projects for which NPV was calculated (1999-2011), the median net present value was $48.5M, and ranged from $3.5M – $773M.

The same logic applies to the IRR metric, the AIRR metric, and the B/C metric. A project has an acceptable economic impact if the IRR on the capital invested exceeds the cost of capital. The Executive Branch Office of Management and Budget stipulates the Government cost of capital for investment projects like AES to be 7%. Clearly the AES program’s 81% IRR indicates significant impact. Similarly, a program B/C greater than one has a sufficient impact to go forward. The AES program’s B/C of 29/1 (the present value of benefits/present value of costs) indicates that for every $1 of NIST’s investment, society received $29 dollars’ worth of cost avoidance benefits.

Explanations of the economic impact metrics based on extrapolated estimates (in the third and fourth column of Table ES-1) are discussed in the following report. Although they show significant increases in the economic impact for the AES program, they are nonetheless, as the report explains, lower bound estimates of the true economic impact of the AES program.

The opening paragraph of this executive summary observes that about $40 billion annually is used Government-wide to support intramural research. The AES program is an example of such intramural research. A very small part of that $40 billion—about $2.7 million—was spent for the AES program in 2017, close to the inflation-adjusted annual amount ($2.6 million) spent to support the program over the years from 1996 through 2017. Thus, the annual intramural Federal R&D spending for 2017 is almost fifteen thousand times greater than NIST’s annual spending for the AES program. Now consider the success of the AES program in that context and ask, “What would the annual return on investment be if the collection of Federally funded intramural R&D investments were as productive as the AES program?” For simplicity, imagine the expenditure of $40 billion annually had continued indefinitely. The annual yield on that $40 billion would have to be $72 billion to justify an internal rate of return equal to the AES program’s 80%.