In this dissertation, we examine some of the important drivers of quality and compliance performance. The context of this dissertation is pharmaceutical industry, where quality and compliance is crucial for public health and thus is heavily regulated by the U.S. Food and Drug Administration (FDA). Despite the critical importance, however, the failure incidents in this industry have been steadily increasing over time, a trend that runs counter to those observed in other regulated industries such as airline and railroads. Yet, research concerning this important phenomenon in the pharmaceutical industry is limited. We consider three broad categories of quality and compliance drivers – pharmaceutical policy, manufacturing strategy and organizational learning. By investigating how the interplays between these categories of factors influence pharmaceutical quality and compliance performance, this dissertation not only advances the knowledge in the literature, but also sheds light on the increasing failure trend in this industry.
Two standalone essays in this dissertation examine pharmaceutical quality and compliance performance at different levels: 1) at the drug level, where quality performance is measured by serious drug recalls, and 2) at the inspection level, where compliance performance is measured by violations of current Good Manufacturing Practice (cGMP) regulations as determined by the FDA during the plant inspection, respectively.
In the first essay (Chapter 2), we compare the quality risk of generic drugs, whose approval and use has been facilitated by the government to reduce healthcare costs, against that of the corresponding original brand-name drugs. Then, we further examine whether manufacturing drugs in less-advanced economies, an increasingly prevalent manufacturing strategy in the pharmaceutical industry, influences the generic drug-quality risk relationships. Based on a large-scale, drug-level panel dataset, we find that there is no significant difference in the frequency of serious recalls between generic drugs and their brand-name counterparts overall. However, generic drug recall risk becomes significantly higher than that of the original counterparts, when they are manufactured in less-advanced economies. Our post-hoc analysis indicates that learning-by-doing, as measured by drug age, significantly mitigates the generic drug quality risk relative to the corresponding original drugs, but that internal manufacturing and learning from failure experience does not. These findings have implications for policy makers and managers. For managers, there is potentially a greater cost for quality and compliance associated with manufacturing generics in less-advanced economies. For policy makers, there should be an increased focus on generics manufacturing in less-advanced economies; and there may be value to requiring transparency regarding drug manufacturing location.
In the second essay (Chapter 3), we examine vicarious learning from the FDA warning letters – one of the means for the agency to disseminate failure information to the industry. More specifically, we investigate if the probability that the focal plant violates a cGMP regulation as determined by the FDA during the plant inspection decreases as other plants receive more warning letters citing that specific cGMP regulation. We also explore if such learning becomes more pronounced among the plants that are similar in one of two dimensions – geographic location and manufacturing processes. Our analyses, based on the FDA’s public facing databases, reveal that there does not exist significant industry-wide learning or learning among the plants that have similar manufacturing processes. We do find, however, that the focal plant learns from the warning letters issued to the other plants located in geographical proximity. These findings imply that, in the pharmaceutical industry context, the FDA’s role as a central source of broadcast transmission of knowledge – one of the salient knowledge diffusion mechanism established in the vicarious learning literature – may not be sufficient alone to induce compliance-enhancing learning.
Overall, this dissertation enhances the understanding of how some of the important pharmaceutical policy-related factors (i.e., promoting market competition via low-cost generic drugs, disseminating failure information via warning letters), together with manufacturing strategy (i.e., manufacturing location, outsourcing) and organizational learning (i.e., learning from own and others’ experience), affect quality and compliance performance. The findings of this dissertation not only advance the aforementioned literature bases but also provide valuable implications to policy makers and managers. Our findings may be generalizable to other industry contexts, especially where quality is not easily observable and thus compliance is difficult to measure. This dissertation raises questions for future research about the quality and compliance implications of other pharmaceutical policies, such as Generic Drug User Fee Act (GDUFA) or Drug Supply Chain Security Act (DSCSA), whose effect may be contingent on some organizational or supply chain characteristics.