Over the last 150 years, the U.S. chemical industry has contributed significantly to both the national and global economy (Aftalion 2001; Arora et al. 1998). The industry’s contributions to economic growth, employment, and improvements in life expectancy, health, and living conditions in Western-style societies are widely recognized (NRC 1992; Spitz 2003). The industry’s products are ubiquitous; in roughly the last 50 years, synthetic chemicals have become integrated into nearly all industrial processes and commercial products and now constitute the primary material base of society (Geiser 2001).

The scale of chemical production is correspondingly enormous: Data from the TSCA Inventory Update Rule (IUR) show that the United States produced or imported about 15 trillion pounds of chemical substances during the 2002 reporting cycle, or about 42 billion pounds per day (U.S. EPA 2005). For the 2005 reporting period, chemical manufacturers reported producing or importing about 27 trillion pounds of 6,200 chemicals at more than 25,000 pounds per site per year, or about 74 billion pounds per day. The IUR data include substances used in industrial processes and products and do not include fuels, pesticide products, pharmaceuticals, or food products. There is no clear explanation for the 80% increase in volume between 2002 and 2005, and the U.S. EPA has obscured the categories of the products in which each chemical is used (U.S. EPA 2009). The TSCA inventory now lists about 83,000 substances that have been for sale in the United States at some point since the inventory was first published in 1979 (U.S. EPA 2008a). Of these, approximately 62,000 were in commercial use at the time TSCA was passed in 1976, and about 20,000 new substances have entered commercial use since that time (U.S. GAO 2005a).

Global chemical production is projected to continue growing—about 3% per year, with a doubling rate of 24 years, rapidly outpacing the rate of global population growth (Figure 2) [American Chemistry Council 2003; Organisation for Economic Cooperation and Development (OECD) 2001; United Nations 2004]. This growth will distribute globally both the benefits and the health and environmental consequences of industrial chemical technologies.

The cases noted above illustrate the expanding health and environmental problems attributable to an antiquated U.S. chemicals policy that has failed to keep pace with developments in either green chemistry or the environmental health sciences. Given the size of the chemical enterprise, the extent to which it is woven into the fabric of society, and the backlog of unexamined chemicals, a new approach is needed that does not rely on resource-intensive, chemical-by-chemical risk assessments in which government, at great public expense, bears the burden of proof. An integrated chemicals policy is needed that enables identification and prioritization of chemicals of concern and employs market and regulatory tools sufficient to motivate investment by industry in the design and production of safer chemicals and materials, based on the principles of green chemistry. Making this transition will require a chemicals policy that departs markedly from the federal policies of the last 30 years, of which TSCA is emblematic.

When it was passed in 1976, TSCA grandfathered the 62,000 chemical substances that were in commercial circulation at that time; that is, except on a case-by-case basis, chemical producers were not required to generate and disclose any information about the uses or hazard traits of these products (U.S. GAO 1994). In essence, this body of existing chemicals was assumed to be safe unless the EPA could prove otherwise. To gather the necessary hazard and exposure data from producers, however, TSCA §4 required the U.S. EPA to establish, on a chemical-by-chemical basis, a) that a substance may present an unreasonable risk to human health or the environment, or b) that there is either significant human exposure potential or substantial quantities of the chemical are produced, imported, and released into the environment (TSCA 1976). This created a logical paralysis for the U.S. EPA: To systematically assess risks of existing chemicals, the EPA needed hazard and exposure data that producers were under no obligation to provide, unless the EPA could first show that such an unreasonable risk might in fact exist or that high exposures were already occurring. This has proved all but paralyzing for the U.S. EPA. In the first 15 years under TSCA, the agency was able to review the risks of about 1,200 (2%) of the 62,000 existing chemicals, despite the fact that the agency estimated that about 16,000 (26%) were potentially of concern based on their production volume and chemical properties (U.S. GAO 1994). TSCA §8(e) does require that chemical manufacturers, processors, and distributors notify the U.S. EPA of any new or unpublished chemical hazard information, and the EPA receives about 300 such submissions each year (U.S. EPA 2003). Perversely, however, this requirement creates a disincentive for manufacturers to voluntarily investigate the hazard properties of their products.

Although the TSCA inventory has grown to about 83,000 substances, the body of 62,000 existing chemicals still constitutes nearly all chemicals in commercial use in the United States. In the 2002 reporting year, 3,000 high production volume (HPV) chemicals—those produced or imported at more than one million pounds per year—made up > 99% by volume of the 15 trillion pounds of chemicals in commerce. Although these HPV chemicals constitute only one-third of existing chemicals by count, their high volume raises concern about the lack of basic hazard information on existing chemicals. Although the U.S. EPA’s New Chemical Program requires some minimal data on chemicals introduced since 1976, these chemicals make up < 1% of the production volume of the substances under the jurisdiction of TSCA (U.S. EPA 2005).

In tacit acknowledgement of its constraints, the EPA has turned to voluntary initiatives to close the data gap on existing chemicals. The HPV Challenge (U.S. EPA 2008b) is one such effort, begun in 1997 as an effort to gather screening-level data on HPV chemicals. It has been substantially limited, however, by late, incomplete, and poor- quality data submissions by chemical producers (Denison 2007; Denison and Florini 2004). In 2008, the U.S. EPA announced the Chemical Assessment and Management Program, in which the agency plans to conduct risk-based prioritizations for about 6,750 chemicals (U.S. EPA 2008a). These assessments, however, rely on incomplete data and information from TSCA inventory updates, much of which is obscured by trade secret claims and a lack of transparent assessment methods (Denison 2008).

Given that the U.S. EPA bears the burden of proof, the agency is further constrained by the level of evidence required to take regulatory action. To regulate a chemical, TSCA §6 requires the U.S. EPA to provide substantial evidence of all of the following conditions: The chemical presents or will present an unreasonable risk to health and the environment; the benefits of regulation outweigh both the costs to industry of the regulation and the lost economic and social value of the product; the U.S. EPA has chosen the least burdensome means of addressing the source of unreasonable risk; and that no other statute could adequately address the risk (TSCA 1976).

Under this evidentiary burden, the U.S. EPA has been able, since 1976, to use its formal rule-making authority to partially regulate five existing chemicals (or chemical classes): polychlorinated biphenyls (PCBs), chlorofluoro carbons, dioxins, asbestos, and hexavalent chromium (U.S. GAO 1994). Of these, an amendment by Congress to TSCA required regulation of PCBs, and the U.S. EPA’s asbestos regulation, promulgated after the agency spent 10 years building its case, was overturned in its most significant aspects by the 5th Circuit Court of Appeals, which concluded that the U.S. EPA had failed to meet its burdens of proof under TSCA (Goldman 2002).

TSCA does grant the U.S. EPA relatively more authority to regulate new chemicals introduced since 1976, as well as new uses of existing chemicals. Producers must submit premanufacturing notices (PMNs) before marketing a new chemical. However, there is no required minimum data set beyond information already in the possession of the producer at the time they file the PMN. This has created a disincentive for manufacturers to investigate their products’ potential hazards because, although known hazards must be reported, ignorance of hazard is not penalized. Not surprisingly, an EPA evaluation found that 85% of PMNs lacked data on chemicals’ health effects, and 67% lacked health or environmental data of any kind (U.S. EPA 2003). Hindered by limited data and the short time period permitted for the agency’s review, the U.S. EPA has taken some form of action on < 10% of the 36,600 chemicals that producers have proposed for commercial use between 1979 and 2004 (U.S. EPA 2007). Once new chemicals have entered commercial use, the U.S. EPA may regulate them only under the standards and burdens it carries for existing chemicals, as described above.

Extensive trade secret claims permitted under TSCA have exacerbated both the data and safety gaps. Although some protection of proprietary information is necessary, in practice the statute’s allowances for confidential business information (CBI) claims have severely limited access to basic information on chemical identity and use. In 2005, the U.S. EPA reported that 95% of PMNs submitted by producers contained some information claimed as confidential (U.S. GAO 2005a). One assessment found that 90% of the CBI claims in PMNs hid the identity of the chemical (U.S. EPA 2003). CBI allowances under TSCA have thus contributed to a pervasive lack of supply-chain transparency about chemical hazards, despite other regulations intended to facilitate hazard communication via Material Safety Data Sheets (Sattler et al. 1997; Occupational Safety and Health Administration 2004; U.S. Government Accountability Office 1991).

By requiring less hazard information for existing substances than for new chemicals (data gap) and by stymieing effective regulation of well-known hazards (safety gap), TSCA has produced a chemicals market that favors existing chemicals over newer and potentially less toxic substances. As a result, the statute provides little motivation for industry investment in green chemistry research and development. The vast majority of chemical products manufactured in the United States rely on technologies developed 40–50 years ago, a fact that led the Council for Chemical Research to call for new technologies that incorporate economical and environmentally safer processes, use less energy, and produce fewer harmful byproducts (Council for Chemical Research 1996). Twelve years after the Council’s Vision 2020 report, the Web sites of the 50 largest U.S. chemical companies all state their commitment to reaching sustainability goals, but their spending on research and development has remained level or declined since about 2000 (Lempert et al. 2003; NRC 2005). A National Research Council report concluded that this trend makes it difficult to advance the science and technology needed to support such sustainability goals (NRC 2005).

Public investment is also critical to the development of green chemistry. The U.S. experience in other sectors illustrates that investment by government, federal laboratories, and research universities helps spur innovation of new technologies (Block and Keller 2008).

It is a reflection of the priorities of the U.S. chemicals market that, with rare exception, students can earn an undergraduate or graduate degree in chemistry at universities throughout the United States without demonstrating an understanding of the principles of toxicology, ecotoxicology, or green chemistry (Cone 2008c; NRC 2007). Educational priorities have largely matched those of the chemicals market, such that the hazard traits of a substance, for example, are undervalued in the chemistry classroom relative to chemical function, price, and performance. Without a policy strategy that will favor green chemistry in the market, and without a corresponding research and educational effort, the United States risks lagging behind the European Union and other industrialized regions in the scientific and technical development of green chemistry.