Unintended Consequences of Products Liability: Evidence from the Pharmaceutical Market*

Abstract

In a complex economy, production is vertical and crosses jurisdictional lines. Goods are often produced by an upstream national or global firm and improved or distributed by local firms downstream. In this context, heightened products liability may have unintended consequences on product sales and consumer safety. Conventional wisdom holds that an increase in tort liability on the upstream firm will cause that firm to (weakly) increase investment in safety or disclosure. However, this may fail in the real-world, where upstream firms operate in many jurisdictions, so that the actions of a single jurisdiction may not be significant enough to influence upstream firm behavior. Even worse, if liability is shared between upstream and downstream firms, higher upstream liability may mechanically decrease liability of the downstream distributor and encourage more reckless behavior by the downstream firm. In this manner, higher upstream liability may perversely increase the sales of a risky good. We demonstrate this phenomenon in the context of the pharmaceutical market. We show that higher products liability on upstream pharmaceutical manufacturers reduces the liability faced by downstream doctors, who respond by prescribing more drugs than before. Eric Helland Claremont McKenna College Department of Economics Claremont, CA 91711 eric.helland@claremontmckenna.edu Darius Lakdawalla Schaeffer Center for Health Policy and Economics University of Southern California 3335 S. Figueroa St, Unit A Los Angeles, CA 90089-7273 and NBER dlakdawa@healthpolicy.usc.edu Anup Malani University of Chicago Law School 1111 E. 60th Street Chicago, IL 60637 and NBER amalani@uchicago.edu Seth A. Seabury Schaeffer Center for Health Policy and Economics University of Southern California 3335 S. Figueroa St, Unit A Los Angeles, CA 90089-7273 seabury@usc.edu A common feature of modern, complex economies is vertical production that crosses jurisdictional lines. Technological development and specialization is associated with vertical chains of production, where upstream firms supply inputs for downstream firms who add value and sell to consumers. Moreover, firms at all levels have grown in scale and scope, and they now often serve many markets across a range of different jurisdictions. This geographic expansion is perhaps even more pronounced for upstream firms, because downstream firms, such as distributors and retailers, tend to be local or at least retain a well-defined local presence. To remain relevant and effective, tort rules and other legal structures need to account for the interactions among firms in a vertical chain of production. To some extent, tort law has successfully incorporated the reality of vertical production. In the 1800s, a doctrine called “privity” prevented individuals from suing upstream firms for injuries from products of downstream firms. Cases such as MacPherson v. Buick Motor Company (N.Y. 1916) and Smith v. Peerless Glass Co (N.Y. 1932) abandoned the doctrine of privity and allowed consumers to sue firms further upstream (Prosser, 1960). Indeed, contemporary products litigation is now characterized by suits against several firms in the vertical chain of production and has become quantitatively significant.1 Overall torts liability grew four times faster than the overall economic growth rate between 1930 and 1994 (Sturgis, 1995). By 2009, total payments in products liability suits alone amounted to $248.1 billion, or 1.74% of U.S. GDP (Towers Watson, 2010). In health care, suits against doctors amount to 1-2% of physician expenditures (Mello, Chandra, Gawande, & Studdert, 2010), and suits against drug companies amount to 2.26% of all drug expenditures.2 However, an important way in which tort law has lagged the modern economy is in the persistence of local, rather than national or global tort rules. States set tort rules, even though firms may produce for national and even global markets. This has encouraged beggar-thy-neighbor policies by states who may have incentives to shift liability from local downstream defendants to upstream national defendants lacking a local presence (Krauss, 2002). For instance, 22 states have reduced the products liability local retailers face but not the liability that upstream manufacturers face (Shepherd, 2012). Nearly 30 states have caps on total or non-economic damages that physicians face in medical 1 For example, plaintiffs sue both the manufacturer of the car whose tire burst and the manufacturer of the tire (e.g., In re Bridgestone/Firestone, Inc., S.D. Ind. 2003), the home builder that used contaminated materials and the maker of those materials (e.g., In re Chinese Manufactured Drywall Products, E.D. La. 2010), the grocer that used spoiled food and the company that supplied the grocer with the spoiled food (e.g., cases against retailers and farmers implicated in the 2006 E. coli outbreak), and the doctor that prescribed a drug as well as the company that produced it (e.g.,Wyeth v. Levine, U.S. 2009) This last example is also the topic of the empirical application in this paper. 2 This estimate is derived from all settlement special items reported in the income statement. For pharmaceutical companies this represents provisions to alter reserves for litigation and settlement. For other companies the amount would include insurance payments from the firms general liability policy but pharmaceutical firms do not typically have insurance against loses in litigation. As such the sum of the settlement special items represents unexpected payments in litigation. Although some of this litigation is likely not related to product liability the vast majority of loses in excess of reserves is likely major product liability cases—a fact reflected by disclosures in the 10k statements. We sum the total special reserves incurred from 2002-2008 and divide this by total sales over the same period to get the ratio of 2.26%.