The US government classifies information to protect it from falling into the wrong hands. Secretly gathered information on a foreign nation’s nuclear weapons posture, identities of US intelligence agents, or profiles of foreign leaders using secret informants clearly warrant such protection. Still, there is a broad bipartisan consensus that too much information is classified.
The government has tried to reduce the amount of information deemed classified. However, past efforts have essentially amounted to pleas to the individuals with authority for classifying information to do it less. One critically important reason there is so much classified information is that a decision to classify specific information does not entail any direct cost to anyone.
Moreover, although any properly cleared individual can challenge the classification of a specific item of classified information, such challenges occur only rarely. These conditions create a mindset of “better safe than sorry, and when in doubt, classify.” Pleas to classify less will not overcome pro-classification incentives, which include a genuine desire for protecting truly sensitive information, inhibiting public debate on controversial actions or programs, and covering up agency embarrassments.
But budgets matter more to agencies than exhortations, and if an agency incurred a budget penalty every time it classified information, it would probably make fewer such decisions. An agency could be required to budget an overall line item to cover its acts of classification, and an agency that exceeded it would have to explain the overrun to Congress and/or take money from other parts of its budget to cover the excess cost. That would provide significant incentives for the agency to limit the number of its (now-costly) decisions to classify information.
Giving an agency the ability to earn credits for classification of current information by declassifying older classified documents could decrease the overall volume of classified information. But the most powerful measure would be a stipulation that allowed an agency that stayed under its classification budget in a given year to keep the surplus for discretionary spending in the next year, subject to appropriate Congressional approval.
How big should an agency classification budget be? In 2017, the most recent year for which figures are available, the US government made about 49 million decisions to classify information and spent about $18.4 billion on the classification system — an amount that was a few percent of overall national defense expenditures in 2017. Accordingly, the line item should be a few percent of an agency’s budget for activities related to national defense and security.
Dividing $18.4 billion by 49 million decisions to classify corresponds to a cost of about $375 each. (A decision to classify information occurs when a properly authorized person makes a decision to classify a set of words containing information that in his or her judgment is classified.) But the precise dollar figure isn’t as important as the principle that an agency should incur a nontrivial cost for each such decision. Indeed, agencies themselves could assign a reasonable dollar figure for decisions to classify. An unrealistically low figure would acknowledge that the value of classifying information was low and cast serious doubt on the claim that disclosure of the information affected would damage national security. An unrealistically high figure would lead to a premature exhaustion of their ability to classify information.
One possible method of implementing this approach uses the presence of a classification mark for every paragraph in a classified document that contains classified information — the total number of words in all classified paragraphs can be a measure of how much classified information the document contains. In practice, subtleties arise, such as the fact that information can contain multiple classification levels or that documents contain images.
Herbert Lin is senior research scholar and fellow at Stanford University.