Back in 2014, the New York Times reported that the Central Intelligence Agency had been paying the office of then-President of Afghanistan Hamid Karzai tens of millions of dollars in cash for more than a decade. Afghan officials termed these payments “ghost money,” a convenient term that I adopt here—though some might simply call it bribery. This case was hardly unique. Indeed, the practice of engaging in state-sponsored bribery in the interests of national security appears to be a longstanding and global one: Over last half-century or more, the CIA has reportedly made cash payments to heads of state from Angola to Zaire in exchange for favors.

U.S. officials have defended this controversial practice. One former CIA operations officer even went so far as to say that state-sponsored bribery serves a productive role in the anticorruption fight: where the CIA is asked “to monitor the level of corruption in a place like Afghanistan,” “it only makes sense that U.S. operatives would have to talk to, and if necessary, bribe those involved in the corruption to find out what is going on.”

Yet even if one sets aside the question of whether ghost money itself presents the same normative concerns as regular bribery by private parties (an issue previously discussed on this blog), ghost money raises more problems than it solves for the anticorruption fight. In particular, the U.S. practice of making ghost money payments in places like Afghanistan likely has three significant adverse collateral consequences:

• First, ghost money payments are likely to be recycled into other forms of corruption. The New York Times described U.S. payments to the Afghan president as going largely “to paying off warlords and politicians.” In other words, ghost money lets recipients buy support from clients. Leaders like Karzai might engage in clientelism regardless of whether they received ghost money, and in theory, ghost money might play a neutral or even beneficial role if it simply replaced other, more troubling sources of influence, such as misappropriation of private funds, or reliance on personal wealth or threats. But in practice it’s unlikely that ghost money merely displaces other sources of leverage for political patrons. Rather, ghost money likely adds to the pot of resources available to those at the top. In doing so, it can quickly transform into ordinary private corruption.
• Second, ghost money undermines the anticorruption messaging, and personal credibility, of U.S. diplomats and other advocates of anticorruption reforms. Karzai and others who receive ghost money are likely to think U.S. diplomats calling for anticorruption reforms are doing so with a wink and a nod—or else are clueless about their own government’s practices. The public revelation of ghost money payments can have a similarly damaging effect on the credibility of official U.S. statements criticizing corruption, and taint even legitimate foreign aid by tapping into suspicions about the motives of foreign NGOs and other entities. Indeed, the revelation of ghost money payments can jeopardize other forms of financial assistance to a country, even including funds meant for anticorruption purposes. For instance, after learning of the payments to Karzai, Senator Bob Corker blocked $75 million in aid for the Afghan government, saying that “further commitment of taxpayer funds at this time toward such an incoherent governance strategy would not be in our national interest.”
• Third, ghost money payments can weaken the enforcement of laws against corporate bribery, such as the Foreign Corrupt Practices Act (FCPA). For one thing, because sponsor states reportedly use private corporations to provide cover when they pay ghost money, sponsor states may be willing to exempt those corporations from normal accounting requirements, or even to weaken such requirements altogether. This practice resulted in the “national security exception” to the FCPA’s bookkeeping and internal accounting requirements, which allows U.S.-listed issuers to avoid the requirements “[w]ith respect to matters concerning the national security of the United States” for a person “acting in cooperation with the head of any Federal department or agency responsible for such matters….” A pair of professors has conjectured that in at least two cases—involving Lockheed Martin and Westinghouse Corporation—CIA involvement served as a plausible explanation for the absence of FCPA prosecutions, perhaps because the national security exception was invoked. And the practice of ghost money payments might damage FCPA prosecutions by making some defenses available or more plausible—namely, that the government (a) actually authorized a defendant to pay a bribe or (b) that it led the defendant reasonably to believe that she was authorized to pay a bribe (see the Second Circuit’s dicta in United States v. Giffen).

So what should be done? Representative Barbara Lee has introduced a bill three times (see here, here, and here)—most recently, in January 2019—that would “prohibit monetary payments by the Federal Government to employees, officers, and elected officials of foreign countries” for bribery or similar purposes. That bill includes a national security exception allowing payments “if the President certifies to Congress” that complying with the statute “would harm the national security of the United States or members of the Armed Forces.”

The Lee bill is a step in the right direction. But ending the inquiry at whether a payment serves some national security purpose (the answer to this broad question will almost always be “yes”) misses the real question: does a payment’s national security purpose outweigh the collateral consequences, including anticorruption ones? Any efforts to rein in ghost money should balance national security goals, such as collecting useful intelligence and building stable governments in war zones, against other concerns. If this is the relevant substantive question, then the second question is an institutional one: who should decide (and review) the cost-benefit analysis?

The Lee bill would be more effective were it to (a) require the executive branch, in the context of each payment or set of related payments, to analyze the likelihood and significance of collateral consequences and justify the expected national security benefits in light of those risks; (b) require a politically accountable executive official to certify this analysis; and (c) allow an institution with at least some degree of political influence, budgeting authority, or legally binding power to review the executive’s analysis. The reviewing institution could range from a congressional committee to an administrative tribunal; the review itself from de novo to highly deferential; and the consequences of a flawed determination from professional embarrassment to denial of funding

The ideal contours of review merit greater discussion and debate. But that debate should happen, and it should follow the basic principles outlined above: that ghost money has collateral consequences for national interests, that those consequences should be weighed against national security interests, and that executive determinations of that balance should be able to withstand at least some form of independent review.