Work crews sort through hazardous materials dropped off for disposal in Washington. (Lois Raimondo/The Washington Post)

Joseph Allen is an assistant professor at Harvard University’s T.H. Chan School of Public Health, where he directs the Healthy Buildings program and acts as the faculty adviser to the Harvard Healthy Building Materials Academy.

When new parents see the words “BPA-free” on a baby bottle or sippy cup, they are meant to assume that the product is safe. This may well not be the case — quite to the contrary.

In fact, in some cases, hormone-disrupting BPA, or bisphenol-A, has simply been swapped for a similar chemical — BPS, or bisphenol-S — that may well pose even greater dangers to child health. In this way, manufacturers have done an end run around on the much-publicized dangers of BPA without addressing the underlying problem.

Even more disturbing: What’s happened with baby products is the tip of the iceberg. It points to a phenomenon known in the world of public health as “regrettable substitution” — the cynical replacement of one harmful chemical by another equally or more harmful in a never-ending game being played with our health.

Over the past four decades, such swapping has become commonplace. Some have compared the practice to a big game of whack-a-mole: Every time one chemical is knocked out, another takes its place. Remember DDT, the pesticide used through the 1970s in the United States? DDT is associated with delayed mental development in children and, when banned, was replaced with organophosphate pesticides, another class of chemicals that can also interfere with a child’s developing brain. (DDT itself was a “safe” replacement for toxic lead arsenate, which had been used as a pesticide since the 1800s.)

Flame-retardant chemicals used in furniture tell a similar story. So-called PBBs were used beginning in the 1970s. PBBs are teratogenic, meaning that they interfere with embryo development, and eventually were banned, only to be replaced by PBDEs, which have a similar chemical structure (one molecule difference). PBDEs, we later discovered, are associated with health effects such as decreased sperm production, failure of testicles to descend in babies, hormone disruption and, recent research I led found, higher rates of thyroid disease. Can you guess what happened with PBDEs? They were subsequently banned in some states beginning in 2003, only to be replaced by other flame-retardant chemicals, including one called “tris.” Tris, it turns out, is carcinogenic and is being replaced. TPP, another replacement, is associated with decreases in sperm concentration. Progress!

But the whack-a-mole analogy fails to capture the scale of the problem. We might better think of the mythical Lernaean Hydra, where each head that is cut off regenerates multiple times. Heard of “C8,” the chemical used to make stain-resistant clothing, furniture and nonstick pans? It was used for more than 50 years before scientists found a probable link to high cholesterol, kidney cancer and testicular cancer. C8, aptly named because it has an eight-carbon chemical backbone, was banned 10 years ago and, as of 2011, there were a stunning 268 variations in the same chemical family — the Hydra problem. (Disclosure: I am serving as an expert witness in litigation involving C8.)

In all of these cases, and more — from nail polish to flavorings in e-cigarettes — the chemical replacements can be just different enough that they are treated as distinct from a regulatory and market standpoint.

How have we allowed this to happen? Most people assume, incorrectly, that chemicals used in products are thoroughly tested by the Environmental Protection Agency or another federal agency for health safety before they are allowed to be used in commerce. Except for pesticides and pharmaceuticals, this is not the case. Many people are shocked to learn that even asbestos, infamous for causing lung cancer, has never been officially banned by the EPA. It was only last month added to the list of high-priority chemicals for the EPA to evaluate. In the United States, our chemical policy largely follows the approach of our legal system: innocent until proven guilty.

Changes made this year to the woefully outdated 1976 Toxic Substances Control Act left many with the impression that loopholes are being plugged, but that is far from true. While welcome, the new law simply requires the EPA to prioritize chemicals for safety testing. The EPA will review a minimum of 20 chemicals at any one time. It could take decades to evaluate the 80,000 chemicals already in commerce that have yet to be tested, let alone the 2,000 new chemicals introduced each year. And it still treats each chemical individually, continuing the saga in which similar, but slightly different, chemicals can be regrettably substituted.

Perhaps most concerning is that even this woefully slow process can still be slowed further. President-elect Donald Trump has indicated his disdain for the EPA and suggested massive budget cuts for the agency. Any reduction in funding or staffing, or a deprioritization of chemical issues, could derail even these incremental approaches to testing chemicals for safety.

Innocent until proven guilty may be the right starting point for criminal justice, but it is disastrous chemical policy. We need to recognize regrettable substitution for what it is: repeated substitution of toxic chemicals with equally toxic chemicals in a dangerous experiment to which none of us knowingly signed on. The Trump administration should recognize the significant health and economic benefits of a strong, fully implemented chemical policy.