“The story of America in the twenty-first century is the story of chosen scarcities.” This is the message from Abundance, the new book by Ezra Klein and Derek Thompson. In it, they argue that the left of the last few decades has created scarcities by being overly focused on subsidies:
Progressivism’s promises and policies, for decades, were built around giving people money, or money-like vouchers, to go out and buy something that the market was producing but that the poor could not afford.
In America, this includes subsidizing healthcare with the Affordable Care Act, food with food stamps, housing with subsidized housing, and many other things with the minimum wage.
Klein and Thompson don’t oppose these policies, but they argue that demand-side solutions are at best only part of the solution. After all, subsidies don’t lead to more stuff. As they say, “if you subsidize demand for something that is scarce, you’ll raise prices or force rationing”. And so here we are with unaffordable or inadequate access to housing, healthcare, public transit, education, childcare, and a bunch of other stuff.
Crucially, the Democrats in America and the Liberals in Canada—that is, the left—did this. California, which has had a Democrat governor since 2011, has 50 percent of America’s unsheltered homeless population. During a decade in power, the Liberal Party under Justin Trudeau did precisely nothing to meaningfully address Canada’s housing problem. What they did try, including taxes on foreign buyers and vacant homes, attempted to lower demand (and made no difference).
For healthcare, American health insurance premiums have risen by more than 300 percent since 1999. In Canada, it’s wait times that have grown.
According to Klein and Thompson, the way out of this is for progressives to care about building and inventing stuff. The left has been too focused on subsidizing demand at the expense of increasing supply, which too often leads to debates about demand-side issues instead of addressing underlying supply constraints. For instance, the evidence is clear that we can’t make housing affordable with rent control. We actually have to build houses. Even in your backyard (or near it).1
I care about this because vibrant cities are central to a thriving society, and cities can’t be vibrant if people can’t afford to live in them. But, as I’ve said a bunch around here, some bioethics issues are actually downstream of the broader problems Klein and Thompson are pointing to. If housing is playing a role in people choosing MAID, then we should increase the supply of housing, not ban people from accessing MAID.
There are also lessons for bioethicists, which tend to focus on demand-side solutions. Here’s one by way of a comparison. In Abundance, Klein and Thompson describe the multi-decade failure to build high-speed rail in California. (Klein also discusses this in this recent article in the Times.) Californians first passed a bill to begin construction in 2008. The planned route was supposed to connect San Francisco with Los Angeles and was to be completed in 2020, but here in 2025, there are zero miles of working track. One major source of delay is the lengthy environmental review process, which began in 2012 and still isn’t finished. It’s the reams and reams of red tape, not engineering problems, that has stopped the lines from being built.
Now compare this to Scott Alexander’s description of trying to run a small study to see if a screening test for bipolar disorder is effective. Alexander, a practicing psychiatrist, was a resident at the time, and he recounts how the institutional review board at his hospital made conducting the study so difficult that he and his colleagues eventually gave up after multiple years of trying. The bipolar screening tool was already being used by psychiatrists on patients, but he wasn’t able to conduct a study on the same patients to see if it worked. Additionally, many of the IRB’s requirements—including that psychiatric patients, who aren’t typically given pens for safety reasons, be required to sign their consent documents in pen—were understandable but misguided.
Building train lines and conducting medical research are clearly different, but, in these cases, their failures have a common cause. As with environmental review, each IRB requirement might be justifiable on its own, but inflexible rules can produce bad results of the sort Alexander describes. And, as with environmental reviews, the result of piling on more and more rules is years of delays or, in these cases, no train and no research.
This is bad because medical research improves and saves lives. People can be harmed in the research process, but people are also harmed when research that could have saved them gets held up. As Alexander points out, an overly bureaucratic system also means that only the largest pharmaceutical companies will have the resources to navigate the system, while researchers at smaller institutions or citizen scientists—the ones who aren’t motivated by profit—are left out.
Just as Klein and Thompson aren’t calling for the end of government oversight, my point here isn’t that we should do away with ethics boards. Researchers sometimes do bad stuff, and we need oversight to prevent that from happening. Likewise, the Department of Government Efficiency shouldn’t be the model here. Firing people arbitrarily, as Musk’s team is doing, doesn’t improve efficiency and it won’t prevent researchers from doing bad stuff. Instead, it’s just going to slow down research more.
That bioethics has the same blindspot as the left isn’t surprising. Nearly everyone in bioethics is on the left. According to a survey of American bioethicists last year, only 4.8 percent of them identified as conservative. Twenty-eight percent said they were “very liberal” and 47 percent said “liberal”. I suspect Canadian bioethicists are distributed similarly, if not more left. (I also suspect the ‘very liberal’ number would be higher if the researchers had used the term ‘progressive’ instead.)
We need to attend to the supply-side of bioethics.2 For research, that means balancing the protection of human subjects with the goal of producing as much valuable research as possible. To do this, we have to take seriously that blocking research in the spirit of erring on the side of caution causes problems. We want both: the goal is a lot of ethical research, so we need to be aware of how the demandingness of policies can interfere with the quantity of research that gets produced.
This doesn’t require us to abandon our role as exposers and addressers of unethical conduct. I once saw an ethicist block a patient from being wheeled to surgery because there were serious ethical issues with the consent process. The ethicist stood in the hallway on the other side of the bed, refusing to move. I’m sure she didn’t yell “You shall not pass!”, but she might as well have. She was in the right, and it remains one of the coolest things I’ve ever seen.
We also can’t eliminate scarcity altogether. Bioethics is in the scarcity business. As I’ve written, there aren’t enough organs to go around, so we need to figure out how to distribute them. When the Covid vaccines were developed in record time, we still had to determine who should get access to them first. When the hospitals were near capacity during the pandemic, we needed triage protocols. These resource-allocation questions aren’t going away, and we shouldn’t ignore them.
Still, the story of bioethics in the twenty-first century is the story of chosen scarcities. It usually isn’t bioethicists making these choices alone, but we play a role. Bioethicists have spent far more time figuring out how to allocate the organs than we have than trying to figure out how to increase the supply of organs. IRBs spend far more time adding procedural steps than reviewing their processes to see if all the steps are required, and some bioethicists have spent a lot of time opposing challenge trials, thereby decreasing the supply of consenting research subjects.
Supply-side bioethics means thinking about the ways that supply limits good options, then finding ways to address the shortage. This requires approaching problems differently. Demand-side problems aren’t going away, but, at least in some cases, the problem is scarcity, and scarcity is a choice.
[1] The Liberal Party of Ontario’s economic illiteracy is evident by its proposal to introduce province-wide rent control. Thankfully, the federal Liberals under Mark Carney are focusing on increasing supply. But it shouldn’t take choosing a literal economist as prime minister to figure this out.
Some bioethicists are already doing this. For example, Paula Chidwick’s team at William Osler created the Checklist for Ethical and Legal Obligations (ChELO), which focuses on improving the processes and structures of medical decision making. There’s also increasing awareness that the supply of physicians, especially family doctors, is a problem.
Very interesting article. Enhancing scarcity is definitely a policy choice. For example rent controls are not really about the economics of rental housing, they are about its political appearance. Governments imposing rent controls are making a statement: we are on your side, not your landlord's. That's visible, while the deterrent effect on the supply of new or additional rental housing is invisible. That visible/invisible division is probably also valid for bioethics.
What's the right balance between innovation and regulation? Conservatives often are in favor of deregulation, trusting the market to sort things out, whereas progressives argue regulation provides necessary safeguards.
The ChatGPT copyright infringement debacle is a good example. AI moguls have argued that honoring copyright would have been prohibitively expensive and so they're "moving fast and breaking things." AI, it seems, is ushering a new era of how we think about copyright. This is still in development so who knows how this will pan out, but it seems like the scales might tip in favor of innovation and away from regulation, much to the detriment of anyone who has produced copyrighted material. The irony is that the innovation here won't necessarily produce anything better - at least in the realm of the humanities, where we see a lot of AI slop.