Of Algorithms, Algometry, and Others: Pain Measurement & The Quantification of Distrust

By Frank Pasquale, Professor of Law, University of Maryland Carey School of Law

Many thanks to Amanda for the opportunity to post as a guest in this symposium. I was thinking more about neuroethics half a decade ago, and my scholarly agenda has, since then, focused mainly on algorithms, automation, and health IT. But there is an important common thread: The unintended consequences of technology. With that in mind, I want to discuss a context where the measurement of pain (algometry?) might be further algorithmatized or systematized, and if so, who will be helped, who will be harmed, and what individual and social phenomena we may miss as we focus on new and compelling pictures.

Some hope that better pain measurement will make legal disability or damages determinations more scientific. Identifying a brain-based correlate for pain that otherwise lacks a clearly medically-determinable cause might help deserving claimants win recognition for their suffering as disabling. But the history of “rationalizing” disability and welfare determinations is not encouraging. Such steps have often been used to exclude individuals from entitlements, on flimsy grounds of widespread shirking. In other words, a push toward measurement is more often a cover for putting a suspect class through additional hurdles than it is toward finding and helping those viewed as deserving.

Of Disability, Malingering, and Interpersonal Comparisons of Disutility (read on for more)

The current political context is not encouraging, either. Economist David Autor has chronicled the rise in disability claims, and conservative advocates have seized on his work to try to demonstrate that there are more undeserving claimants than ever. (NPR’s scaremongering on the topic provoked an unprecedented rebuke from eight former Social Security Commissioners.) It is easy to imagine a future variation on SSA’s “grid system” involving higher “objective” bars based on algometry. Another pretender to the “science of suffering,” the “fake bad test,” has been used to root out “malingerers” by analyzing their responses to a series of test questions. Critics claim it is unfair to many with real, but unusual or unpredictable, suffering.

The Wall Street Journal, in an article on changes to the disability determination process, eagerly reminded readers that “many workers can continue working well into their 60s and 70s in quite productive positions.” That could be encouraging in the context of reporting about age discrimination. But it’s more ominous here, given the paper’s jeremiads about “overspending” on the disabled.

Pain and disability also have very different salience in different societies – and depending on how we perceive our own society. A starving village may expect even its most beleaguered residents to help farm a crop; some more affluent societies are now experimenting with unconditional basic incomes. Majorities in the United States, and some other advanced economies, consider their societies to be closer to the precarious, impoverished village than to the secure position they actually inhabit; thus they favor stringent conditions for and rationing of benefits. This is a testament to the power of propaganda to make people in affluent societies insecure, not to our actual and considerable ability to provide the disadvantaged.

Given these social Darwinist background political understandings—or imaginaries of work and leisure, desert and entitlement—it feels almost inevitable that algometry, as it is translated into policy contexts, will create some “gold standard” benchmarks of suffering, while marginalizing others. This needn’t be intentional—indeed, it may arise out of purely well-intentioned work to identify the most deserving. Yet, this is simply a predictable side-effect of commensuration, as policymakers in Oregon discovered when they pursued explicit ranking of services. There could easily be “a general tendency for automated decisions to favor those who belong to the statistically dominant groups,” in part because “by definition that there is always proportionately less data available about minorities.” Algometry may simply shift the burden of exclusion from the linguistically to the mathematically unrecognized.

Admittedly, a kinder, more empathetic political environment could lend an entirely different political valence to the project of developing more objective, neurological bases for metrics of pain. Scientists may plead that they have no special responsibility to anticipate misuse of their tools. But we should be careful to contextualize policy based on algometry as part of a larger project of “measurement that yields the possibility for bodies to be lived as fundamentally comparative on the level of…the life of the population and species” (a quote from Mader’s Sleights of Reason). That is all too often a neoliberal governance strategy: to force the social welfare arms of the state to “do more with less,” by justifying cuts in aid to some portion of the population by characterizing them as having failed to meet some statistical standard.

Questioning the Quantitative

Another worry, more for the profession of law and regulation in general, is the implicit valuation of quantification and scientific imaging as intrinsically more verifiable or objective than narrative accounts of human experience. Cost-benefit analysts have enthusiastically propounded their work as a new science of policy, rendering more fair and transparent decisions that were once made on murky grounds. But that ostensibly neutral tool of CBA has been warped over the decades, whether via selective deployment (it can block regulation in the US, but not deregulation), or biased implementation (consider the undervaluation of benefits of financial reform).*

Moreover, there’s an extensive literature on the difficulty of “interpersonal comparisons of utility.” We should also recognize the problem of incommensurability of “interpersonal comparisons of disutility,” too. Narrative accounts of the difficulties faced by an aid recipient are one way of doing justice to the diversity of trying situations that the modern sick and disabled find themselves in. Again, the neoliberal response (already seen in Mathews v. Eldridge) is to emphasize the assumed expense and contestability of qualitative evaluation, relative to quantitative measurement. But in this, as in so many other precincts of power, preferred modes of decisionmaking are infallible because they are final, and not vice versa. Open up the science to contestation, and you will get plenty of challenges.

From a policy perspective, I’d be much more amenable to the “pain-o-meter” project if it were premised on the possibility that far too many people are working with pain, and we need to find ways to identify them, and to (oblige employers and the state to) relieve the conditions that compound that pain. From a professional perspective, I’d suggest to other attorneys (looking at brain-driven algometry for a holy grail of objective determination of pain) that the “grass is always greener”: Technology all too often seems to possess a consistency and authority lacked by systems of mass justice like disability determination, until one begins to press hard on the inevitable gaps between lived experience and its numerical or pictorial representation. As Ricoeur suggests in “Creativity in Language,” words’ indeterminacy enables certain freedoms even as they frustrate plans, consistency, and governability—and the bitter must be taken with the sweet.

*Consider also Doug Kysar’s work challenging cost-benefit analysis, and confidence in quantification generally, or Heinzerling and Ackerman’s Priceless.

Earlier posts in this series:

This post is part of the series on pain, brain imaging, and the law sponsored by the Center for Law, Brain & Behavior at Massachusetts General Hospital, the Petrie-Flom Center, and Harvard University’s Mind/Brain/Behavior Initiative. Contributors participated in the conference Visible Solutions: Now Neuroimaging Helps Law Reenvision Pain. For inquiries, please contact the organizer Amanda C. Pustilnik (@apustilnik on Twitter).

Be Sociable, Share!