How Our Days Became Numbered: Risk and the Rise of the Statistical Individual, by Dan Bouk

In middle school, we had career week. Most professions on display were well known, but one – the subject of a school-wide speech – was not. I recall to this day sitting with a couple hundred others listening as a man explained why we ought to do well in math class: ‘actuaries,’ he described, were not only important business people, they also relied on those mathematical skills liable to bore a 12-year old. Scrolling through lengthy tables, he impressed us with the ability to divine our futures based on readily accessible data. Not only did math seem to offer specialized skills, our teachers were quick to emphasize that this was a well-paid job – not unlike the more familiar doctors and lawyers to which many of us aspired. While I am unsure how many of my preteen peers became actuaries, I suspect at least some now identify as ‘data scientists,’ that rapidly growing profession of our big data present.

Dan Bouk’s wonderful new book is a timely history of the roots of our contemporary situation. He explores how the American life insurance industry transitioned from its late-nineteenth century aspirations of predicting fate to an early-twentieth century effort to master death. It is not only a history of actuarial science, but also a cultural history of capitalism – and a surprisingly gripping tale for an industry that many, preteens included, are liable to find dull. Bouk narrates the story through a cast of characters who traipse through graveyards, assemble massive databases, and investigate corporate malfeasance. Statistics about life and death, he demonstrates, are anything but boring; rather, they are animated by and occasion moral debates about family, race, and the future of the nation.

At the core of actuarial science and thus the life insurance business is a tension between two mathematical tendencies: ‘classing’ and ‘smoothing.’ Classing seeks to slot customers into ever more accurate categories; its apogee would be a perfectly individualized policy whereby no individual pays more or less on account of others. Smoothing represents another ideal: the cultivation of overarching regularities through the statistical aggregation of individuals. Neither pole alone was technically possible nor commercially desirable, but bias in one direction or the other has implications for who has access to life insurance and at what cost. Bouk demonstrates (particularly in chapters one and four) that classing versus smoothing is a central dialectic of calculating lives.

Calculation on this scale, however, requires considerable infrastructure. Life insurance companies developed large-scale databases of applicants and their relevant details. As ledgers proved insufficient, they became early users of index card systems. The clerks who ran these systems relied upon a substantial network of salesmen and doctors who were responsible for gathering the data, yet they were often at cross-purposes. Those interacting with applicants had an incentive to provide viable customers, sometimes to the point of fudging the data. Those maintaining the databases were wary of anything that would undermine algorithmic accuracy. Doctors, too, chafed at the new demands on their professional judgment. Such friction produced new practices and innovations: to curb undesirable applicants, insurers shared data in what amounted to a blacklist. In other work, considerable energy was put into training doctors to see and write in standardized ways, yet the goal was not to produce automatons. Instead, doctors’ professional judgment was enrolled by the insurers to serve as frontline diviners of aches and pains – and what they may herald about the future.

If life insurance is, today, largely unremarkable, that was not always the case. Plenty of folks were understandably uneasy with the arrival of strangers bearing predictions of their death. Bouk demonstrates how such unease was eventually converted into its opposite: being a good family man came to mean providing for your kin in the event of your untimely demise. Life insurance was sold as a moral good, but largely to men. Even when they cared to do so, insurers did not know how to value unwaged labor, so women were more likely to be seen as domestic dependents. This was frontier capitalism, opening new realms to commodification and economization, and its proponents muddled through the epistemological, technological, and cultural work of stabilizing a market.

Not only women were illegible to insurers. African Americans, too, were often excluded because insurers were unable, in post-Civil War America, to use the historical past to make sense of the future. The end of slavery marked too radical of a break in the historical record, and for reasons linking discrimination and statistics, insurers more readily decided simply not to sell to African Americans. The entwinement of life insurance with race is a multifaceted tale, and Bouk tells it well, chronicling activist campaigns against the practice as well as the travails of African American insurance entrepreneurs. Some aspects strike a contemporary reader as bizarre: one Prudential statistician received considerable attention for a book predicting not a hopeful new future for America’s liberated black population, but rather the inevitable extinction of the race. Other aspects continue to bedevil our own age: the statistics available to turn-of-the-century actuaries reflected historical injustices, and there was the threat that any system reliant upon those statistics would end up reproducing such inequality.

The science and infrastructure enabling the large life insurers’ business created a new mass commodity in the form of risk. The American public was enrolled through popular media and the proliferation of scales for weighing oneself. In the terrific fifth chapter, Bouk shows that the annual medical checkup – a biopolitics so normalized as to be unremarkable for us – has its roots in this culture of quantification. Irving Fischer, today mostly known as a pioneering economist, was also a vocal social reformer. Instead of merely predicting death, Fischer proposed what he called a ‘modern conception of death’ whereby insurers and others would control the future through changes in the present. In its uglier versions, this included eugenicist plans to remove criminals and other degenerates from the nation. Through the Life Extension Institute, he also advocated investment in the infrastructures of public hygiene and increased personal reform (neither of which are immune from discriminatory ends, it should be noted). Fischer and others, armed with statistics and an optimistic sense of forestalling death, convinced the American public of a national crisis where others had not previously seen one.

Eventually, the type of registration and enumeration pioneered by life insurance companies would inform the New Deal’s creation of Social Security, and in turn, members would have a lifelong identifying number for reincorporation into commercial insurance systems. If actuaries and their peers had done much to number our days, it was a legibility significantly furthered by the government.

And today, Bouk reminds readers in the epilogue, it is private industry, the government, and willing consumers who are energetically quantifying an unprecedented swath of life – much of it headed to insurance firms eventually. For example, in 2014, Microsoft and American Family Insurance partnered to promote ‘home automation.’ The proliferation of sensors within homes and on our bodies allows for not only better tracking of behavior but, as a sequel to How Are Days Became Numbered might demonstrate, the normalization of certain practices through incentives, penalties, and today’s favorite form of regulation, ‘nudges.’ The feedback of data generated through smart watches and fridges is funneled through Silicon Valley and insurers and back at us (and often our peers). As Dow Schüll (forthcoming) found among its proponents, such ‘data for life’ is reworking and exemplifying ideas of self-regulation.

Nor is this only a phenomenon of the wealthy world. Where I work, in Kenya, telcos and banks are busily generating data from mobile phone use in order to offer loans to subscribers. Like the subjects of Bouk book, these businesses are attempting to determine reliable and profitable commodity forms – less insurance (for now) than credit – that rely upon enumeration and fixing the relationship between smoothing and classing. Thus far, though, these businesses have not received the sort of popular or political attention that emerged with the American life insurance industry, meaning we know much less about their methods and implicit biases. Bouk’s book, however, will be a crucial resource if and when this happens.