Tyler Cowen offers an optimistic theory about the longevity of the surveillance state:
Let’s say that everything is known about everybody, or can be known with some effort. The people who have the most to lose are powerful people who have committed some wrongdoing, or who have done something which can be presented as wrongdoing, whether or not it is. Derelicts with poor credit ratings should, in relative terms, flourish or at least hold steady at the margin.
It is not obvious that the President, Congress, and Supreme Court should welcome such an arrangement. Nor should top business elites. More power is given to the NSA, or to those who can access NSA and related sources, and how many interest groups favor that?
Therein lies a chance for reform.
I’m not as sanguine, and I don’t think Mancur Olson would have been either. The security state is a perfect, if subtle, example of Olson’s “dispersed costs, concentrated benefits” thesis. A classic application might be America’s Farm Bill (which costs each of us cents each year, and gives large agricultural many millions, providing incentive to lobby).
Here’s a place to start for this thought experiment. How much would you pay to avoid government surveillance. This isn’t a realistic heuristic, at least to the extent that those willing to pay the most are also most likely to be the ones we want to target. But it’s sound enough to build a public choice logic. I wouldn’t pay more than $10 a year, and even that small change is a reflection of my distaste for the program rather than any rational calculation. Frankly, I just don’t care if all my information is read by a jobless bureaucrat at the NSA.
On the other hand, a whopping 98% of Booz Allen Hamilton’s revenue comes from the security state. Here’s the nonlinearity, the big “business interests” Cowen talks about perhaps have a Bayesian prior of p = 0.005 that government surveillance will affect their profits negatively – or the probability that information regarding the company surfaces, that it is detrimental, that government would be willing to use it against them, and that it actually effects long-run profits. A long chain. (My point here isn’t restricted to a binary consideration of profits, but makes the abstraction a little bit simpler). And that’s an overestimate, if you ask me. I do believe businesses powerful enough can evade state scrutiny through a plethora of tools, including encrypted networks. Strong corporations also have a good reason to believe were sensitive material to emerge, they can strong arm the government into compliance.
But Booz’s prior that they will benefit from the military-intelligence-industrial complex (MIIC) is asymptotic to p = 1. A better way to put it is their prior that they will be destroyed without the MIIC is p = 1. And there’s a huge number of companies in the most politically connected corners of the country who will have this prior.
Therefore, for Cowen’s suggestion to ring true, the 200*E[cost of MIIC to big business] > E[benefit to MIIC]. That seems extremely improbable to me unless, as some have suggested, the MIIC directly hurts technology companies in foreign countries.
Also remember, we’re discussing metadata aggregation here. This is far more relevant to understanding political trends, indeed it’s what made the 2012 Obama campaign so technologically robust. It seems very unlikely that specific, detrimental, messages – of, say, an affair – would surface beyond the intestinal gears of whatever mining algorithm NSA uses.
This is, of course, so long as NSA doesn’t target businessmen in particular. But even if lobbies can prevent them from doing that, until Americans decide to place a non-negligible monetary value on their privacy, I don’t see the emergence of a permanent security state.
By the way, we know that Americans don’t trust their government. A recent study showed, however, that people respond to surveys with “cheap talk”, as a means to display political affiliation. What if our indifference towards surveillance is a revealed preference that we do, indeed, trust our government?