How Our Days Became Numbered: Risk and the Rise of the Statistical Individual

At its core, this book is a history of the life insurance industry. It is the two or three layers of meaning surrounding this core that make the book so incredibly rich. It is a history of changes in society, and changes in the mindset of those of us who participate in it.A theme throughout the book is the increasing quantification of risk and measurement in all aspects of life. Life insurers are able to quantify the risk of mortality. They moved pursued that in increasingly fine detail and quantified individual risk factors such as smoking, bodyweight, age and sex. One of the markers that they tracked was race. The book deals quite a bit with African-Americans, and their fight to obtain life insurance at premium rates comparable to those of whites. Death of course involves the question of health, and I was surprised to learn how long ago life insurance companies took an interest in the health of their customers, both out of corporate altruism and the desire not to pay out death benefits.Life insurance is a data-driven business. The nosy insurance men annoyed everybody involved in the process in their quest for data. They asked doctors to fill out health evaluations, lawyers to do credit checks on potential clients, and applicants for life insurance to fill out personal histories. All of them eventually agreed. This is a central point of the book – we have acquiesced to being treated as numbers. The benefits, such as life insurance, in general outweighed the privacy and dignity being given up.The life insurance companies came to control huge pools of money. Their business model is to collect now and pay later, if at all. They make their money by investing the premiums. Self-interest being what it is, they pay themselves handsome salaries and have been known to stiff their customers. The book makes very good use of two 19th-century legal cases to illustrate the moral perils of the business.This book is a history up to about the time of the introduction of Social Security under Roosevelt. It does not address Obama care, Medicare, or other more recent programs.The chapter titles are:ClassingFatalizingWritingSmoothingA Modern Conception of DeathValuing Lives, in Four MovementsFailing the FutureConclusion: Numbering in LayersA brief description of each follows.Classing: Risks are different for different people. Over any period of time – for instance, the next year – an older person is more likely to die than a young one. An overweight person more likely to die than somebody of average weight. A man more likely to die than a woman. Lastly, a black person is more likely to die than a white person. Every life insurance company recognizes different classes of insured people. Age groups, people with pre-existing conditions, people who do not have very good medical histories.In the early days of life insurance just after the Civil War, life insurance companies did not have what we would call good databases to draw on in pricing their products. It was an extremely competitive business, with hundreds of companies. They wrestled a great deal with the problems of classing. They tried everything: family histories, ethnic background, phrenology, occupation, dietary habits and much more. Who should they ensure, and how much should they charge? They by and large did not ensure white housewives. They started out not ensuring working-class people. However, when they exhausted the market of middle-class white men they had to expand. The author, Dan Bouk writes very clearly about how they did it.The insurance regulators had an interest in classing as well. They needed to ensure that the companies that they regulated had enough reserves to pay potential claims. A lot of insurance companies went broke, leaving the policyholders with no recourse. Then, as now, it appears that the regulators were dependent on the companies they regulated for actuarial data.Fatalizing: Life insurance is unique in that everybody is assured of dying sometime. A standard policy will pay off. The focus of the chapter is on taking the mystery out of death, reducing risk from a fuzzy possibility to something quantifiable and therefore marketable. This entails a change of mindset among the potential customers. Quite a bit of the book, among the more fascinating passages, are the ways in which the insurance salesman attack the sensibilities of the potential insured in order to sell the product. This plays to one of Dan Bouk’s great strengths. Throughout the book he tells it as it is, with a straight face. He does not moralize.There are two sides to life insurance. Classing deals with statistical abstractions – how many will die within a group of people with similar characteristics. Fatalizing deals with individuals. Life insurance is sold one policy at a time. The individual in question will either live or die, a positive or a negative financial result. This chapter deals with the individuals.We learn that Southern whites were greater risks than Northerners, as well as that Blacks were greater risks than whites. How did the insurance companies deal with it? Was it fair? They didn't think twice about discriminating against whites, but very interestingly, as early as 1880, there were state laws against discriminating against Blacks.There are some interesting passages on how insurance was sold. Expanding away from middle-class white men, insurance salesman would canvas working-class neighborhoods collecting weekly premiums of five cents. The insurance covered little more than burial expenses, but by this account a proper burial was a very high priority, perhaps the most that these people could look forward to.Writing: This chapter deals with the mechanics of the insurance business. It was a nationwide business, highly data dependent. It is extraordinary what they were able to accomplish with the primitive tools of the 19th century. Bouk borrows from a 1909 pamphlet to describe the lengthy process of accepting an application and qualifying the customer. There were always huge paper files. Card catalogs, rather like those of old libraries, came in use about the same time as the Dewey decimal system and the great age of American libraries funded by robber barons. Other innovations that were new at the time included the typewriter, the pneumatic tube, and the telegraph. All were put to immediate use.There were some totally manual databases such as a database of rejected insurance applicants shared among the insurance companies.Another connection that was made very early was between consumer credit and life insurance. The life insurance companies did a credit check as part of their acceptance process. In this chapter we learn a lot about the history of credit verification. Like life insurance, it existed long before automation.Life insurance companies led the way in standardizing procedures for medical examinations, application processing, credit checking and other things that had to be done routinely. The insurance companies were early to standardize urine checking, and had a lot to do with the introduction of scales into doctors’ offices.There was a conflict of interest between the writing side of the business and the finance side. The investment side wanted as much money under their control as possible, and they would favor relaxed underwriting standards. The policy writers, and the regulators, pushed for tighter controls. There were some celebrated congressional investigations and legal proceedings working through this conflict of interest.Chapter 4: SmoothingSmoothing is the process of ironing out statistical irregularities in the interests of presenting a coherent picture. An insurer could certainly not generalize mortality based on one week's experience. Even a year is a short time.The first example of smoothing that Bouk cites is by the Bank of England. Its reserves rose and fell rather significantly day by day. However, in reporting to parliament, they chose to use trailing three months figures. It presented a much smoother picture.Mutual Insurance companies have to pay dividends to their policyholders when aggregate claims are less than premiums. However, the period over which these are aggregated makes a great deal of difference. One year's experience may not be representative. Policies are written every day of the year; one cannot generalize from the December policies.The issue of smoothing is similar to the issue the banks face in setting reserves, or the corporations do in reporting nonrecurring expenses in their quarterly reports. It is a matter of judgment. The question here is that the actuaries’ judgment often seems to err in favor of corporate management instead of the policyholders.Chapter 5: The modern conception of deathThe life insurance companies always had a vested interest in the health of their customers. It was in their financial interest at the customers not die on them. Some life insurance companies started to offer an annual physical examination as part of the policy. It was good business, and the annual physical has become a standard practice. Others went much further in working with their customers on health issues. One authored a highly popular book entitled "How to Live." They took an interest in their customers smoking and drinking habits.The changing conception of death was that it was something that could be pushed into the more remote future. People should organize their lives in such a way that they were likely to live a long time, rather than simply and fatalistically await the grim reaper.The existence of a large database of health data, and the increasingly complete data from censuses and other sources, led to widespread concern in the first decades of the 20th century about the general level of public health. It started out with some best-selling books, and the fear was fed by the fact that many inductees into the armed services at the time of World War I were found to be deficient for one physical reason or another. There was a great deal of concern about smoking and drinking. Not noted in the book, this was the time that prohibition was passed.The book makes a detour into a discussion of eugenics. The early 20th century was the high point of the eugenics movement, and many in the life insurance business had opinions to offer. They advocated that the best and the strongest – read white Americans of Nordic stock – reproduce copiously, and that other populations be discouraged from reproducing.There is an inconclusive discussion of intelligence testing. This was one of the earliest applications of the science of statistics. The Army applied across-the-board intelligence testing to inductees during World War I. They got it wrong in some particulars. The tests that they used were culturally biased and English dependent, which meant that immigrants did not do as well as the native born. The patterns that they revealed among native-born Americans were more or less accurate. The author does not note as much, but these problems were largely worked out by the 1930s with the introduction of context free intelligence tests such as Raven's progressive matrices. Read Earl Hunt or Arthur Jensen for a good overview of the state-of-the-art in intelligence research.Bouk discusses attempts to put a value on human life. This of course cannot be done, and we still wrestle with the problem today. His discussion of the early dialogues on the subject is interesting. They included a lot of intelligent people, and they surfaced the major measurements and morality issues that we face today.Chapter 6: Valuing Lives, in Four Movements is a discussion of the ways in which a life insurance salesman goes about selling the product. How do you raise awareness in the prospect’s mind that he needs life insurance? How do you overcome their aversion to talking about death? How do you convince him that the price is reasonable? How do you evaluate a wife's value as a homemaker? How much money do you need to leave your survivors in order to live comfortably? Once again, Bouk is admirably evenhanded in making this presentation. Although it is done from the perspective of a salesman, he keeps his distance and lets you, the reader, evaluate the arguments.Chapter 7: Failing the FutureThere is an interesting discussion on the black life insurance industry that grew up because mainstream companies would not ensure black lives. I note that whereas some black companies, Black entertainment television and Johnson products come to mind, have been exceptionally successful, the black owned life insurance companies were not. It turns out, although Bouk explains it somewhat obliquely, that the actuarial basis for the mainstream companies not to write insurance was well-founded. This raises an interesting question of social justice of which Bouk is aware, but which he chooses not to moralize on.My overall conclusion:This is a five-star book without a doubt. It covers most of the interesting issues of a very vibrant period in American history. It is extraordinarily well researched, with more pages dedicated to footnotes and bibliography than to the text itself.One shortcoming that I note is that he does not fully identify some of the personalities in the book. He mentions Karl Pearson without noting that Pearson was one of the founders of the field of statistics. He talks about Franz Boas as the mentor of Louis Dublin, without mentioning that Boas is the father of American anthropology. Boas was also a relatively strong leftist, and as such opposed the use of ethnic stereotyping and classing in the insurance business.The insurance business always needed more data than they could process. I would've appreciated the author’s making a stronger case for the ties between what the industry needed and what science and technology made possible. Although statistical data was collected throughout the period covered, Karl Pearson and Ronald Fisher developed the science of statistics around the turn of the century. Mechanical adding machines and calculators came in about 1890, and punched card machines were invented for the census of that year. Useful electronic accounting machines were not widely used until the 1930s. Bouk credits IBM with having provided them for the startup of Social Security. The rest of the story is that that was probably the most important single step on IBM's path to becoming the computer powerhouse of the 1960s and 1970s.Bouk presents and challenges a lot of conventional wisdom. I would propose that he challenge a couple of others. First, there is a question of whether money alone can replace a life. Many lives that have been shattered by tragedy turn out to be quite productive. The loss of a breadwinner does not necessarily doom his children to a life of poverty, despite what the insurance salesman may say. There is the question of the payout. If it is all at once, the beneficiary may suffer the lottery-winner syndrome and fritter away the money. On the other hand, if it is structured as an annuity, inflation may eat it away. On an even more global scale, all policies are denominated in national currencies, fiat currencies, which are under stress at this writing. The idea that one can safeguard against all risk is simply a misconception. Bouk, in his discussion of Social Security, does a great job with its beginnings. A discussion of its endings, and the false security that many would claim it provides, might be appropriate.

A well written history. It's expansive but not meandering, pointed but not moralizing. Bouk anchors a dense and technical subject – life insurance – with fascinating human stories and controversies that resonate today. The author's passion for the subject matter and literary sensibility ("Frederick Hoffman spent days upon days riding trains" introduces one chapter) makes the reader want to come along for the ride. Me being me, my favorite sections were grounded by the composer Charles Ives (surprise! he was an insurance guy!) and the writer Richard Wright (surprise! him too!) and a couple delicious little bits I learned about the history of IQ testing (note to self: find out more) and the picture-perfect moment in 1936 where Social Security numbers were first assigned ("the biggest bookkeeping job in history"). Speaking of pictures, the photos are fantastic, most of all that to-die-for Dorothea Lange photo on the cover. Because I've read this book, I now know that the man on the cover is 27 year old Thomas Cave. Finally, Lange's poignant photograph provides a surprising and suitably complex structure for the conclusion.

The history of insurance is much more fascinating than one would expect from what is usually considered a dull subject. This is really the story about how data slowly came to rule our lives, how we went from tabulating deaths from gravestones to modern big data. Fascinating, easy to read and highly recommended

Brilliant, engrossing history of life insurance industry, its beginnings in assessing risk, the folly of measuring heads to assess longevity, various ways prejudice played its part in development of our move to score, rate and predict lives. Beautifully told through stories and main industry players layered with intellectual history. Must read especially for those interested in today's scored society!

I had expected a history on insurance statistics to be boring. This was an unexpected delight to read.The overlap between insurers’ aspirations to collect and use demographic data and labor bureaucrats’ aspirations is fascinating. There are interesting parallels with contemporary debates about big data, which both insurers and big government want to use. Also the point that in the 1890s, this statistical endeavor tended to operate below the radar, and was not fully understood by either company management or politicians until the 1905 Armstrong investigation; compare the opacity of big data methods today. Also ties between the eugenics movement and insurance executives; see today’s enthusiasm for insurers to get their hands on genetic data.A rich and fascinating book.

Original and thought-provoking history of the problem of applying probabilistic risk to individuals. More interesting to general readers than that sounds, probably. Well researched, well written, clever.