Blogs & Comment

Does the punishment really fit SNC-Lavalin’s alleged crimes?

It depends on what our goal of punishment is in the first place

RCMP officer in SNC-Lavalin lobby

An RCMP officer stands in the lobby of SNC-Lavalin’s Montreal headquarters in April 2012. (Graham Hughes/CP)

News that Canadian engineering giant SCN-Lavalin is facing new criminal charges caused a stir last week, not least because of the penalties that are on the table. The fraud and corruption charges spring from the company’s dealings in Libya. And if convicted, in addition to any other penalties handed out, the company could be barred for 10 years from competing for contracts with the Government of Canada—a category of contracts that make up an important element of the company’s income.

Jacques Daoust, economy minister for Quebec (where SNC-Lavalin is based) thinks that rule is too harsh, and was quoted as saying, “In 10 years a company might not be the same. Everything can change. And in the case of SNC-Lavalin, they’ve decided to make drastic changes already over a short time….”

But the appropriateness of the punishment depends in part on what your goal is in punishing in the first place.

Some will say that we need harsh penalties as a deterrent. That is, if we punish wrongdoing—including corporate wrongdoing—wrongdoers should be less likely to repeat their offences. Additionally, other potential wrongdoers are likely to think twice before stepping out of line. But so far criminologists have been unable to find support for the claim that deterrence actually works. It defies our intuitive understanding of human motivation, but we simply don’t have clear evidence (despite having looked for it) that the threat of punishments deters corporate crime.

Another possibility is that punishment is intended to inspire reform. This is presumably what Daoust has in mind. Having already suffered significant legal penalties, SNC-Lavalin has engaged in a process of reforming itself, making internal changes that (hopefully) will change the organization’s character and make it less likely to offend in future. Will it work? Who knows? But reforming a corporation may be more plausible than reforming a human criminal: after all, you can literally re-engineer a corporation to change the way it makes decisions, but you we don’t know how to do that to a human (and if we could it likely wouldn’t be ethical). And if, as Daoust seems to imply, the key reforms have already been implemented, a prohibition on federal contracts would be baldly punitive, a form of punishment for its own sake.

And to be sure, some will cheer for purely punitive actions. Denunciation—the community, through legal mechanisms, effectively shouting “No!” in the face of wrongdoing—is another of the “classical” justifications for punishment. But in engaging in such denunciation, the Canadian government might be cutting off its nose to spite its face, as the saying goes. Because by barring SNC-Lavalin for competing for contracts, just for the pure punitive joy of it, the government would also thereby be cutting itself off from a large, competent, Canadian engineering contractor.

This leaves one possibility, one rationale that would make sense of debarring SNC-Lavalin, in addition to whatever other penalties the company and its officers might suffer. The other classical rationale for punishment—a rationale usually reserved for explaining the value of incarceration—is an interest in removing offenders from settings in which they would be able to repeat their crimes. Criminologists refer to this as incapacitation. When you put the child molester in jail, he no longer has access to children and hence can no longer molest them. Put him away for 20 years, and those are 20 years during which he will molest zero children. By parallel, if you forbid a company from competing for government contracts for 10 years, those are 10 years during which they will engage in precisely zero instances of attempted bribery or other acts of corruption, at least with regard to bribery or corruption of Canadian officials.