When Safety Measures Make Us Unsafe

How everything from seatbelts to bank deposit insurance can backfire

Foolproof: Why Safety Can Be Dangerous and How Danger Makes Us Safe, by Greg Ip, Little, Brown, 326 pages, $28

Sometimes our efforts to be safe have the opposite effect. Bike helmets may seduce riders into taking chances they otherwise would not. So they die. Better to think riding a bike is really dangerous: That leads to more caution, and more lives saved. Same with snow tires—having them lessens anxiety, and presto, careless maneuvers are more likely.

In Foolproof, reporter Greg Ip of The Wall Street Journal takes up many examples of unintended effects. Seat belts, antibiotics, river dams, anti-lock car brakes, fire prevention, saving for a rainy day—all good things that, I fear to say, have at least the prospect of built-in danger.

But some safety measures do work, a lot of times dramatically or at least pretty well. The book's subtitle tells us that safety "can be" dangerous, not that it will be dangerous. This leads to the inconvenient necessity of rational discrimination on a case-by-case basis. Kids driving without seat belts on a Saturday night are a self-destroying menace in a way middle-age women on a Tuesday morning are not. We need, as Ip declares (and delivers), to examine the relevant "history and evidence with an open mind." Even when much of the story is well-known, Foolproof gives us further details that clear up old questions—and sometimes, alas, raise new ones.

Fire prevention is a good place to start. Smokey the Bear had a myopic view of the health of forests—no fires, no way, no day. Some plant species, however, need fire to reproduce; it's part of their nature. Mature trees, Ip explains, survive forest fires because their crowns are above the intense heat churned up below. What does them in are stands of adjacent young trees that provide a ladder for flames to climb to their crowns. Regularly occurring, and thus smaller-scale, fires would have destroyed those young trees. Such fires would also have taken away the heavy kindling that otherwise accumulates on the forest floor. In the longer term, and contrary to Smokey, nobody can prevent forest fires. It is often good to let nature run its course.

Another charge against Smokey: "His campaigns against fires lull people into building houses where they should not go, as do all the paraphernalia on standby to stamp them out. Among other troubles, this shifts costs to insurance companies (who are slow to get the message and raise premiums), to public agencies that deploy personnel and equipment, and to outfits like the Federal Emergency Management Agency that are supposed to follow up with shelter. Ever balanced, Ip agrees that we need to fight fires when there is a direct threat to homes, businesses, or lives. But Smokey needs some culling.

Ip also takes on football helmets. They were an obvious fix for head injuries on the field, long suffered not just by professional athletes but by athletic schoolboys. The problem: Players adjusted to the apparent safety by using their heads as battering rams against opposing players. Head injuries went down, a near-term gain, but injuries were deflected to other body parts. The U.S. saw a threefold rise in permanent quadriplegics and a fourfold increase in broken necks.

But learning does happen. The National Collegiate Athletic Association came up with a ban on "spearing," essentially the practice of using the head to deliberately punish the opponent. As a result, Ip reports, "spinal injuries fell dramatically," so a lot of the safety benefit did survive.

Ip thus refuses to embrace what has been called the "Jevons paradox," named for the English scientist—William Jevons—who advanced the overall concept around 1865. The Jevons paradox asserts, as an almost Newtonian law, that any gain in the security direction will be offset by a loss in the opposite direction. Critics often use it to argue against all kinds of do-good interventions. Ip takes the charges seriously, but his open mind leads him to doubt, for precise reasons he cites, that they provide an accurate picture overall. There may be a tendency for safety to be self-defeating, but it is not a rule, not even a rule of thumb.

In another section of the book, Ip considers common accusations against levees and dams. Configured at least in part to mitigate flooding (and also, of course, linked into systems of power production and water storage), they mislead by instilling a false sense of security. They get tied into insurance protection and government certifications that encourage development on properties in the path of the next overflow. And then the overflow itself is made more severe by the supposed protections: Water channeled by levees or held in a dam becomes a ferocious force when overtopping occurs, walls collapse, and water rushes across the "protected" lands. Katrina, to take the famous case, had greatly lessened in force when it made landfall at New Orleans—classified at that point as a Category 1 storm downtown and no higher than a Category 2 in other parts of the metropolis. The destruction came from the storm surge magnified by the city's artificial flood control artifacts.

Again, Ip is measured: He doesn't think levees are always bad. They're OK, he argues, when coupled with strategic overflow zones that make "room for the river."

A long-time reporter for the financial press (at The Economist before he joined The Wall Street Journal), financial meltdowns are Ip's real métier. He has watched the rise and fall of world economies and the efforts made to stave crises off. He traces how, since the early 20th century, measures have been taken to safeguard individual savers as well as the health of economies overall. When government insures bank accounts, most savers are made secure and their institutions have less reason to worry about panics that will empty out their assets. Score one for intervention.

But again, there are unanticipated outcomes. In return for the deposit guarantees, governments restrict banks' lending policies, require minimal cash reserves, and regularly inspect the books. Newer institutions then take up the practices the banks were forced to leave behind, and older ones (think Goldman Sachs) grow dramatically by getting into the business as well. The upstarts also developed, over time, "instruments" not envisaged by anyone: Credit Default Swaps, Basket Default Swaps, the ABX (an asset-backed securities index), and, with resonance of World War I, the word "tranches."

In the great foolishness that followed, loans were approved with little due diligence. There was no motive to be cautious, because the link between borrower and lender got severed—what economists call the "agency problem." The first loaning institution (banks or otherwise) flipped the loan to a different outfit. Thousands of such assets were assembled en masse into still other products that could again be sold off—in chunks and pieces, as investor-customers might desire. Various firms escaped all vulnerability, in effect taking a fee for simply attracting the applicant and processing the paperwork. It all went fine until the housing bubble burst and the "underlying assets" were revealed to be permeated by rot.

Those who invested and those who had provided insurance against failure (such as AIG) were challenged to stay afloat. Some survived by dint of wisdom. Some were bailed out, in another protective act that may have paradoxically encouraged wrongful risk-taking. (As with paying off a homeowner who built in a flood plain, a bank bailout can sustain the behaviors that made the help necessary.) And some failed, taking down a good piece of the world economy in the process.

Banking and finance are big things; helmets are not. One of the creative strengths of Ip's book is the search for similar phenomena across what appear to be unlike circumstances. He is able to see continuities. One is the persistence of individuals and institutions altering their behavior, unwittingly or wittingly, against those measures designed for safety. So there is always the potential for a Jevons-like effect, but there may also be ways that effect can be neutralized.

Another of Ip's analytic threads is the "compositional effect": What is good for one actor is not good at all if a lot of people do the same thing. If only some people in a stadium stand up to better see the field, they will see better; if everyone stands up, nobody sees better, and now they have the collective hurt of not having a seat. Pulling money from a faltering bank makes sense for the first movers; if a lot of people do it, there is ruin for all the rest. Just as there can be bank panics, there can be "food panics," as when a few bad heads of spinach cause a market collapse for all spinach.

Ip takes on a lot but also leaves a lot out, and I don't think it's simply because giving us more would make the book too long. He skirts some important issues, perhaps out of timidity. How would he deal with those opportunists in the finance industry—what solutions might at least draw down the risks? He doesn't say. Nor does he say much about the politicians who incite the fears that panic the country into such unwise safety programs as the war on drugs or the TSA's war on shampoo. And then there's foreign policy. The post-9/11 wars were supposed to create homeland security but have instead made us less secure. Surely they would be relevant here.

It is a waste of Ip's smarts to hold back on these great issues of our time. Foolproof can go further.