I’m with Sam here in wondering whether the solution to this problem is best driven “top down” (i.e. by the government) or “bottoms up” (i.e. by pharmaceutical companies, hospitals, and other private actors). The key question in my mind is whether private actors bear the full cost of stockouts. Most likely, the answer is no: when a stockout causes death (due to unavailability of key drugs), the social cost is much greater than whatever hospitals charge pharmaceutical manufacturers and distributors.
I wonder: would forcing private actors to bear the full cost of stockouts (for instance, by forcing them to purchase life insurance for patients or paying a fee to the government when patients die due to unavailability of key drugs) compel private actors to build a more resilient supply chain on their own, or will government need to take a more prescriptive, top-down approach, as Daniel suggests (such as mandating stockpiles, conducting studies, and encouraging geographical diversification)? Would imposing such fees lead to more efficiency in the supply chain, or would the threat of fees cause excessive risk aversion (similar to the threat of malpractice suits)?
I’m intrigued by the question at the end of your post—“is Sierra Nevada doing enough to inspire others?”—because the competitive landscape you outline shows many players not only failing to inspire others, but in fact actively discouraging others from undertaking more sustainable and resilient practices. For instance, you note that ABInBev is working to stop the distribution of South African hops to other brewers. This raises the question: when is sustainability a good strategy at the industry level, and when is it a good strategy at the firm level? My hypothesis is that this question hinges on firm size: for smaller players, the economies of scale that result in industry adoption of sustainable practices (e.g. making it easier to find newer and more drought resistant hops) likely outweigh the loss of competitive advantage from industry adoption (e.g. the reputational benefits of being the only firm to adopt sustainable practices). For larger players, the opposite is more likely to be true. Of course, the total social benefit is maximized when all firms pursue sustainable practices, but this may be a question more of regulation than of business strategy.
This question reminds me of some others we have encountered in class, such as IKEA’s sustainability case and Cynthia Carroll’s mineworking case. In both instances, the protagonists were taking great efforts to act in a socially responsible manner, but were in an industry that generates massive negative externalities. In IKEA, despite efforts at sustainability, the fundamental business strategy was to sell fashionable, low-quality furniture that would quickly be replaced. In Cynthia Carroll, she undertook great efforts to reduce the ~30 annual fatalities from mining accidents, but ignored the ~3,000 annual fatalities resulting from chronic related to working in a mine. For JBS, the challenge is the same: “sustainability” initiatives are being undertaken at an operational level rather than a strategic level, and no amount of tracking or wastewater improvements will remove the fact that raising beef simply requires many times more resources than other forms of protein. The suggestions listed at the end of the post (e.g. diversifying to other protein sources) begin to promise a large impact, but the upshot of these proposals is that JBS should simply no longer be in the cattle business.
It’s exciting to hear that digitization is making an impact in Indian education; its ability to personalize learning seems well-suited to a country as large and diverse as India, and where class sizes frequently exceed 30 students. However, I’m worried that most of the innovation appears to be technical (i.e. better apps, more interactive videos) rather than organizational (such as changing the way teachers work, schools are structured, and so on).
Working as a consultant to US school districts, I saw two major challenges to the kind of personalization that digital tools promise in India. The first is principled: every student might be different, but they may be less different than is commonly assumed. For instance, there are likely only a few ways to effectively teach the Pythagorean Theorem, even though there are hundreds of different digital tools and approaches for doing so. The second challenge is practical: personalization is simply very difficult, in part because the existing tools are not very good and in part because using them effectively requires a very different style of teaching than most teachers are accustomed to. Just as we saw in Fabritech that the same employees who work well in a job shop may be a bad fit for an assembly line, so too may the same teachers who are great in a typical classroom struggle in transitioning to a digital one.
What both of these challenges have in common is an emphasis on the “people” problems, rather than just the “technical” challenges. I hope that as digitization progresses in India, these challenges will be given center billing, rather than simply focusing on the engineering problems.
The notion that organs are discarded at a 40% rate is quite shocking and saddening. However, I’m curious what percentage of this waste is due to inadequate ability to predict transplant success (the problem that AI would solve). For instance, if some of the waste is due to geographic mismatch in supply and demand, there is little AI could do. (This seems likely, as certain regions have so many fewer organ donors that the average wait for a transplant is 6x longer than in other regions.) Another explanation might relate to poor incentives for doctors: doctors are judged based on the 1-year survival rate of patients they choose to operate on; therefore, patients they chose not to operate on are not incorporated into their accountability metrics. This potentially gives an incentive to reject “good fit” organs for high risk patients. Even if an AI algorithm could correctly identify good matches, so long as the doctor makes the ultimate decision to operate or not, they will be incentivized to not operate on risky patients. In all, this seems like an area that is ripe for AI to have some impact, but in which AI will need to be paired with other reforms, including marketing and governance changes, in order to be effective.