As an American living in London, I wake up every morning and check statistics: the number of positive cases reported the prior day in both the UK and US, the number of deaths, hospitalizations and vaccine doses administered, the percentage of the population fully vaccinated and the number of days until the government promises to re-evaluate the lockdown’s end. These numbers determine when I might see my family again, when I might receive a vaccine or even when I might be able to meet a friend for a much-needed outdoor pint.

Of course, beneath these numbers may lie unspeakable loss to families and communities. Nevertheless, their quantification and continual visualization and dissemination in mass media can also make them feel like talismans, ripped from context, critical reflection and, oftentimes, the lives of real people. Indeed, their dominance in public discourse of the pandemic reflects the encroachment of neopositivist social science on lives and livelihoods in new ways—ways that have crowded out numerous other important considerations. 

Writing in the Atlantic  last week, Robinson Meyer and Alexis C. Madrigal, the organizers of the publication’s COVID Tracking Project, offered a critical reassessment of their brainchild’s impacts. They commented on how their data were often accepted by policymakers as sacrosanct, even though they were lagging indicators, reflecting incomplete testing and reporting, as well as an utter neglect of unquantified concerns. Despite big data’s “preeminent claim on truth”—especially in policymaking circles—Meyer and Madrigal wrote that “data are just another type of information.” They need to be considered alongside other concerns, including more difficult to measure ones like mental health and civil liberties. 

After months of a strict lockdown, where the UK government has taken a highly cautious approach to reopening despite more than 50 percent of adults having now been vaccinated, Meyer and Madrigal’s measured insight came as a necessary balm. But it also furthered my frustration at a recent piece in the Proceedings of the National Academy of Sciences hailing this as “The Golden Age of Social Science.” Anastasia Byalskaya, Marcos Gallo and Colin F. Camerer, three social scientists from CalTech, argue in the paper that “interdisciplinary teams,” harnessing “the explosive growth of available data and computational power” have reoriented social science towards tackling major problems like the Covid-19 pandemic. Stuck inside my apartment having not seen my family for nearly a year, I can’t help but wonder, whose golden age is it really? 

Buyalskaya et. al’s commentary represents more than just another neopositivist shot fired in longstanding ‘science war’ debates. Its authors argument that “large-scale problems will only be solved”^[1] via their framework for big data-driven collaboration rests on problematic assumptions that have become deeply entrenched during the pandemic. Given space constraints, I’ll limit myself to excavating two.

First, though Buyalskaya et al call for interdisciplinary collaboration between anthropologists conducting ethnographies and economists using “math-heavy methods,” the authors insist that it take place on distinctly neopositivist terms, rather than pluralistic ones. The authors not only dismiss non-quantitative knowledge’s role in solving large-scale problems, but also fail to recognize the limitations inherent to transferring knowledge across domains and the potentially corrosive effects of quantification procedures. Ethnographers, for example, immerse themselves among research subjects precisely because this affords them perspectival knowledge that quantitative surveys and social indicators cannot. As anthropologist Sally E. Merry has lucidly argued, quantifying ethnographic insight can corrupt the method’s sensitivity to subjective experience and compromise its core findings. 

The issues associated with neopositivist terms of debate shaping policymaking have become especially clear with the pandemic and its technocratic governance. As social scientists have scrambled to produce generalizable knowledge for policymaking, they have often gravitated towards readily quantifiable metrics (cases, deaths, hospitalizations) in lieu of more subjective ones (human suffering due to disease and isolation) that defy easy measurement. Even those studies oriented towards mental health often rely on dubious counting stats that only crudely capture individual hardship. A recent article in Foreign Policy, for example, cited a survey of approximately 1,500 adults drawing on the 11-item UCLA Loneliness Scale to come to the bold conclusion that “[l]oneliness, if it increased at all [during the pandemic], did so by a minuscule amount.” Similarly, the 2021 World Happiness Report relied on a zero-to-ten ‘ladder scale’ survey across 95 countries that life satisfaction and happiness remained steady during the prior year. Such statements reflect a worrying confidence in mass surveys’ ability to capture individuals’ mental wellbeing during unprecedented times, as well as the generalizability of qualities as subjective as loneliness and happiness. Indeed, the New York Times has reported that many mental health practitioners, upset that researchers and policymakers have shown little empathy for young people’s actual experiences, have advised patients to ignore social distancing rules. 

Second, Buyalskaya et. al’s vision of a social scientific “golden age” employs a problematically narrow vision of theory, neglecting the integral role of critical theorizing that interrogates assumptions and values in reframing policy debate. The relative absence of critical reflections in mainstream policymaking circles on emergency powers’ oversight and the protection of civil liberties during crisis during the pandemic has demonstrated just how dangerous shoe-horning theory into solely testable models can be. In the United States, this issue has intersected dangerously with political polarization. President Trump’s ignorance and incompetence inspired oversimplified media narratives that portrayed the left as ‘trusting science’ and embracing lockdowns, while the right ignored scientific facts and emphasized personal freedoms. A true social scientific golden age would emphasize dialogue between what Robert Cox referred to as critical and problem-solving theory, such that inquiry into emergency powers’ efficacy develops in tandem with reflection on their merits.  

Buyalskaya et. al are far from alone in this sort of problematic big data triumphalism and this post is not meant to unduly pick on them. And I don’t wish to insinuate that quantitative social science hasn’t saved lives during the pandemic—rejecting its successes entirely would only deepen methodological divides.

Nevertheless, the pandemic has exposed even further how mainstream, neopositivist social science’s continued rejection of calls for pluralism has infused policymaking. What were once methodological disagreements have been reframed during the pandemic as ‘science’ fighting back against a straw man enemy of conspiracy theorists and anti-vaxxers. It’s side-lined concerns about global poverty and civil liberties in favour of the singular policy goal of lowering case counts. And, perhaps most worryingly, it’s created a template for draconian emergency powers that can now be wielded without meaningful oversight, as such oversight can be caricatured as ill-informed foot-dragging.

To get past the veneer of shiny computational models and big datasets, social scientists need to not only accept the virtues of pluralism in their disciplines, but also embrace its utility in policymaking. A true golden age of social science—one that can actually inform meaningful policy change—requires cultivating empathizing with individuals’ loneliness, isolation and mental struggles, rather than solely trying to measure their prevalence across a population using an 11-point questionnaire developed in 1978. It means keeping sight of political rights even as circumstances change and taking seriously the insight of mental health practitioners, even though they rely on small-N. Finally, it means critically interrogating the assumptions behind epidemiological models before accepting their predictions as dogma. 

Unfortunately, nuance and policymaking are uneasy bedfellows, especially during times of panic. Governance requires insight that cannot be contained by the cells on an Excel spreadsheet. 

---

^[1] Emphasis added.