Opportunity Costs

If Russia spent as much on intelligence agencies as the United States does—$52.6 billion in 2013, according to the “black budget” published by the Washington Post last August—would it have been able to stop the suicide bombers who killed 31 people in two attacks in Volgograd early this week? Can you solve the problem just by throwing money at it? And how big a problem is it, anyway?

 Russia doesn’t really have that kind of money to spend on “intelligence”, so let’s narrow it down to the $10.6 billion that the US National Security Agency spends each year. Of the sixteen intelligence agencies working for the US government, the NSA is the one that places the most emphasis on its alleged ability to stop terrorist attacks through monitoring everybody’s communications.

Would the NSA’s $10.6 billion, spent in the same way by the Russians, have stopped the Volgograd bombers? We cannot know for sure, any more than we can know if another billion dollars spent in the United States would have stopped the Boston marathon bombers last June. So maybe we should reformulate the question.

A total of 785 people have been killed in terrorist attacks in Russia in the past ten years, and Moscow does not pay for an operation remotely comparable to the NSA. In the US, a total of 26 people were killed by terrorists in the same period. So does this mean that the NSA has saved 759 American lives in the past decade?

Probably not. Russia has a far worse terrorism problem than the United States, because some 6 million citizens, living in the Muslim-majority republics of the northern Caucasus, belong to various ethnic groups who see themselves as living under Russian occupation. The United States has no comparable domestic groups, and its ferocious border controls make it very hard for foreign-based terrorists to slip into the country.

There was one exception, twelve years ago, when foreign terrorists did manage to get into the United States and carry out an attack. However, the 9/11 attackers were using a brand new technique. Such innovations are very rare, and are only a surprise the first time. No subsequent terrorist attack, in the US or anywhere else, has been remotely as ambitious.

The NSA has certainly not prevented ten 9/11s in the past decade; it’s very unlikely to have prevented even one. But let us accept, for the sake of the argument, that the NSA’s activities have really saved 759 American lives in the past decade. In fact, let’s round it up to 1,000 lives, to make the calculations easier.

That would mean that over the past decade the NSA has spent around $100 billion to save 1,000 American lives. That works out at $10 million per life saved (on the heroic assumption that without the NSA the American terrorism problem would have been even worse than the Russian).

Economists talk about “opportunity cost”: when you spend the money on one thing, you are foregoing whatever benefits you might have got from spending it on something else. Are there other ways of spending that $100 billion that would save more than a thousand American lives?

Consider spending some of it on better pre- and post-natal care for poor Americans. Just a billion dollars a year—an extra $250 per baby—would enable the US to get its infant mortality rate down below Cuba’s, maybe even as low as Portugal or South Korea. Over ten years, that would be 60,000 more American kids who lived to grow up.

Or take highways. Highway engineers can estimate how many people will die each year on a given stretch of highway fairly accurately. It depends on the width and surface of the road, how many sharp curves and blind hills there are, whether there are guard rails, etc. All those things depend on how much money you have to spend on that stretch of highway.

Around 34,000 Americans died on the roads in 2012. Another $5 billion a year, spent on making highways safer, would probably reduce that toll by an extra thousand people each year. Over ten years, it would save around another 60,000 lives.

That’s 120,000 lives saved, and there’s still $4 billion a year left to spend on other life-saving improvements. You almost certainly end up saving at least 150,000 American lives with your $100 billion investment. That’s at least 150 times better than your return on investing the money in the NSA—and we haven’t yet even considered the cost in alienated allies and violated civil rights of giving the NSA all that money.

Unfortunately, Americans dying in infancy or on the highways don’t make headlines, whereas victims of terrorism do. Politically, their lives are much more important, and so that’s where the money goes. Indeed, even making calculations of this sort about the relative value we assign to human lives is thought to be in poor taste.

Never mind. As Herman Kahn, the dean of American nuclear strategists, said when people criticised him for making cold-blooded estimates of how many millions of Americans would be killed as a result of various different US strategies for fighting a nuclear war: “Would you prefer a nice, warm mistake?”