We've told this story in Governing before, but it makes the point so well that I hope you'll indulge my telling it one more time:

There's a common pesticide called Atrazine that's used by farmers in many of the grain fields of the Midwest. Atrazine has seeped into the water supply of several Ohio cities, including the largest, Columbus. That's a problem, because it's a known carcinogen, proven to cause cancer in laboratory rats. Over the years, Columbus has had to check its water constantly to make sure that levels of the chemical do not exceed U.S. Environmental Protection Agency health standards.

It's not a danger to be taken lightly. On the other hand, the more you read about Atrazine, the more questions you begin to have about the whole matter. It turns out, as with many other chemicals, that the amount an ordinary person would consume in a lifetime falls far short of the amount that made the rats sick--in fact, it falls short by a factor of thousands. The Columbus health department, doing its best to make an honest judgment, calculated a few years ago that Atrazine and other pesticides in the local water could be expected to cause the death of a Columbus resident once every 208 years.

I don't bring this up to ridicule the EPA. A death every two centuries is still a death. It's a legitimate moral and political issue. And its significance stretches beyond the confines of environmental policy. Once you start thinking about this, it leads you to ponder the role that the concept of safety plays in the whole universe of public policy.

The Atrazine story somehow dredged up from the recesses of my mind an old Arrid deodorant commercial from the 1950s. "Why be half safe?" the announcer asks in his oiliest, most pompous voice. "Use Arrid--and be sure."

It was brilliant. Let's face it: You never really know whether you have an offensive odor or not. You took a shower this morning, but maybe that wasn't good enough. Nobody's going to tell you. When it comes to stinking up the office, any degree of doubt is too much. Not only is half-safe dangerous; 99 percent isn't exactly reassuring, either. You want an iron-clad guarantee.

The problem of body odor, if I may spend one more paragraph on it, seems a pretty good proxy for the way we do things in the environmental field. We have an Arrid Extra-Dry policy. When something's this important, we don't like to stop short of certainty. The only problem is, certainty can't be achieved.

Most Americans are at least vaguely familiar with the Delaney amendment, the 1959 provision in federal law that all but prohibited the use of any substance proven to cause cancer, regardless of the amounts that might be required to trigger the disease. Although it is no longer on the books, Delaney created a mindset and we are still living with its consequences.

Not all of them pertain literally to the things we eat and drink. Tom Arrandale, our environment columnist, to whom I am indebted for the Atrazine story, has been writing for years about our all-or-nothing policy toward brownfields--the old manufacturing sites whose soil and water are polluted from years of industrial use. As most of the nation's key environmental laws were written in the 1970s and early '80s, a brownfield isn't really fit for re-use until it has been restored to its original pristine condition. In order for an old chemical factory to be redeveloped, the land around it has to be so clean that a child could scoop up a handful of dirt, swallow it and not get any sicker than he would if he did the same thing on a neighborhood playground.

Well, as Tom has pointed out many times, thousands of industrial sites will never be that clean again, no matter how much money is spent to fix them up. There's always a chance that someone could wander onto the property, ingest something toxic and die as a result. But how often might this happen? Once every 208 years? I suspect less often than that. In the meantime, the property is unavailable for a whole range of commercial and industrial activities that are relatively safe to perform even with a lot of dangerous stuff still in the ground. Relatively safe--not 100 percent. If we want 100 percent, we abandon the site for decades, maybe centuries, until all the toxic residue has disappeared. But as we know, there are negative economic consequences to leaving that land unproductive. It's a question of just how safe we have to be.

In the past few years, American governments at all levels have been moving away from the Arrid Extra-Dry position and toward more of a compromise on environmental issues. More of them are talking about risk assessment--the systematic analysis of just how dangerous a particular chemical agent or industrial process is, and how many people would accept being less than 100 percent safe. There are risk- assessment provisions in the 1996 federal Safe Drinking Water Act, and efforts are under way to apply similar language to the issue of clean air.

I hadn't meant to go on so long with these particular examples. What I really want to do is make the larger point that all these troubling questions of safety and partial safety are relevant to lots of subjects that lie outside the boundaries of environmental science.

Here's one from the most mundane corner of everyday life: Have you ever driven through a small town, parked in an angled space on the main street or the town square, noticed how easy it is compared to parallel parking and wondered why more places don't do this?

Well, okay, probably you haven't. I have, though. I also know the answer to the question. The answer is that traffic engineers will tell you that angle parking isn't safe enough. People will sometimes back out of an angled space and right into an oncoming car. How often? Well, more often than if they parallel parked. After all, the engineers will remind you, this is not just a matter of convenience. There are human lives involved here. So the AASHTO Green Book, the Bible of American highway design, issues a stern warning: "Angle parking presents special problems." It is dangerous and should be used only in "special cases."

Traffic engineers have lots of rules like that. They will tell you that there shouldn't be any parking at all on busy arterials. They will insist that one-way streets are safer than streets that allow traffic in both directions. They will say that, in general, the wider a road is, the better it is. And they will be glad to supply statistics to back up the point. The more of these rules a community follows, the closer it can come to being 100 percent safe.

Of course, there are some other values that get trashed in the quest for ultimate safety. Local governments that stick to the Green Book end up turning their major streets into ugly, noisy, oversized speedways, utterly inhospitable to anyone who wants to use them for walking or even standing on the sidewalk. Whether they are in fact safer is a subject of debate. But let's assume for the moment that they are safer, that they do save at least a few lives over the course of time. Are they worth it? Does it make sense to destroy the aesthetic quality of public spaces in order to avoid an unspecified amount of physical harm?

That's not an easy one. If we're talking about 10 fatalities a year in a city of 100,000 people, it's one thing. If it's one death every 208 years, as in the case of Columbus and Atrazine, then I think most people would describe the rules as excessive. In the end, it becomes a game of risk assessment, a little like the one the environmentalists play. But risk assessment itself represents progress. It's a lot more sensible than simply drawing a line in the dust and saying, "Safety first! No compromises and no other priorities." There have to be other priorities.

The urban scholar Roberta Brandes Gratz, who has carefully documented many of the unfortunate results of safety extremism in local planning, points out that they are not limited to traffic engineering. She cites, for example, fire codes that require a heavy fire wall separating retail space on the ground floor and apartments on the upper floor of two-story commercial buildings. Few older buildings have such a wall, and constructing one adds thousands of dollars to the cost of each upstairs residential unit. Recent studies suggest that similar protection is provided by the use of a sprinkler system, which most of the buildings do have. The net result of the rigid rule is massive amounts of upper-floor space that stands unused, because landlords can't afford to meet the code.

Are these buildings a little safer with fire walls? Probably so. And the possibility of fire is so frightening that most rational people are willing to go to great lengths--even redundant lengths--in order to prevent it. Are the rules justified in this particular case? It's really a judgment call, isn't it?

And perhaps that's the ultimate point to be made on this whole subject. Unlike users of Arrid Extra-Dry, we can never really be sure that we have completely protected ourselves from the risks of any activity. We are stuck with being half-safe, whether we like it or not. The one thing we can know for sure is that we are never going to get to 100 percent, and that a one-dimensional attempt to get there is going to make us more and more uncomfortable the harder we try.

"Salus populi suprema lex," the Romans used to say. "The people's safety is the highest law." They were right. It is the highest law. It just isn't the only law.