Abstract
In recent years, scholars in the social sciences and humanities have turned their attention to how the rise of digital technologies is reshaping political life in contemporary society. Here, we analyze this issue by distinguishing between two classification technologies typical of pre-digital and digital eras that differently constitute the relationship between individuals and groups. In class-based systems, characteristic of the pre-digital era, one’s status as an individual is gained through membership in a group in which salient social identities are shared in common with other group members. In attribute-based systems, characteristic of the digital era, one’s status as an individual is determined by virtue of possession of a set of attributes that need not be shared with others. We argue that differences between these two types of classification technologies have important implications for how persons attach (or fail to attach) to groups, and therefore what kinds of political mobilization are possible. We illustrate this argument by examining contention over the use of gender as a variable in the pricing of risk in insurance and credit – two markets in which individuals directly encounter class-based and attribute-based systems of classification, respectively.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
We are simplifying here since both “nominal” (type or kind) and “ordinal” (score) classifications exist in analog and digital forms (see Fourcade & Johns, 2020: 814). Nevertheless, there is a clear affinity between Fourcade’s nominal and ordinal forms of classification and pre-digital and digital technologies, respectively, that we exploit in our analysis below.
This is not to suggest that such group memberships no longer matter in the allocation of social goods in societies in which algorithms govern decision-making, but they matter in ways that are not directly visible in the score, as we discuss below.
Note that in this article we are translating a broader set of concerns about how the digital is reshaping political life into a somewhat narrower concern with how algorithms – particularly those involving scoring technologies – are changing the potential for political mobilization.
We are indebted to Jonah Stuart Brundage for this formulation.
It is noteworthy in this regard that Michèle Lamont and Virág Molnar (2002: 188) conclude their expansive survey of the study of social boundaries by calling for “a more elaborate phenomenology of group classification” that would identify “how individuals think of themselves as equivalent and similar to, or compatible with, others.”
We take our title from Marcel Mauss’s (1985) essay on the “category of the person.” But rather than examining the category of the person in broad anthropological terms, investigating the meaning of personhood in diverse societies and across long spans of time, here we consider the person of the category more narrowly by examining how forms of pre-digital and digital classification differently configure the possibilities for personhood – especially as expressed through political mobilization – in contemporary American society.
We return to and elaborate on these observations in the conclusion of the article.
Here is another reason for preferring the language of “attribute-based” over “algorithmic” in describing these systems of classification. The scoring technologies we examine in this article represent only one type of algorithm, albeit an algorithm that is increasingly prevalent and influential in a wide variety of decision-making contexts. But many other computational techniques used to aid decision-making are available (e.g., MacCormick, 2012), and we do not presume that all these techniques operate on group formation in the same way we describe for scoring technologies. Again, our aim in this article is not to theorize the implications of the rise of algorithms in general for the shape of political life, but rather to examine the impact of one particularly important algorithmic technology that intersects a broad domain of critically important social decisions (see O’Neil, 2016).
We might for this reason be tempted to refer to attribute-based systems as “individualized,” but this is misleading: these are both methods of constituting individuals, albeit with different intuitions about what it is to be a person, as we elaborate below.
Simon overlooks the possibility that actuarial classifications might in rare instances constitute actual groups founded on common experiences and shared interests. See Nan Hunter’s (2008) intriguing exploration of the risk pools organized by group health insurance plans as a potential foundation for democratic deliberation in the workplace.
On the contrary, Fourcade (2016: 187) suggests that scoring technologies are a primary vector by which group identities are reinscribed as moral differences: “Since credit scores are ‘blind’ to categorical differences in their design, any categorical difference in credit behavior that surfaces appears to be rooted in the relative ‘merit’ (or moral ‘nature’) not simply of individuals but of categorically different populations, as if one could identify some essential, moral difference between them.”
We provide some evidence for this below by examining the National Organization for Women’s efforts to organize opposition to insurance pricing by mobilizing “low-mileage” drivers as the constituency harmed by insurers’ classification practices.
We are indebted to conversations with Nathan Wilmers and Sasha Killewald for the discussion in this paragraph.
The key issue here seems to be the number of observations needed to generate reliable estimates of loss. Insurers’ reliance on cross-classified data may result in a small number of observations per cell, even when insurance databases are large (Chang & Fairley, 1978: 27). This means that insurers will be reluctant to discard variables on which they have accumulated observations, favoring variables already in use in a classification system (see Abraham, 1986: 78–9; Casey et al., 1976: 108). In addition to such technical constraints, insurers’ use of class-based methodologies imposes what we might consider cultural limits on the selection of pricing variables, as well. Risk plans must be socially acceptable: individuals who purchase insurance are typically very sensitive to being grouped with others who are “like” them as a basic indication of fairness (Ferriera et al., 1978: 131). Since these sensitivities are shared by state regulators who must approve risk classification schemes, insurers are not able to substitute variables freely in response to efficiency, cost, or other considerations. Creditors are considerably less constrained with regard to both technical and cultural aspects of pricing systems, as we discuss below (see footnote 46).
Notably, multivariate statistical techniques were adopted much earlier in credit than in insurance. Durand’s (1941) early research relied on discriminant analysis, and as the practice of credit scoring developed this method of analysis continued to be used, supplemented by other techniques such as linear regression, logistic regression, and decision trees (Hand & Henley, 1997: 524). In insurance, by contrast, multivariate techniques were slow to develop, and were still considered experimental (and even treated as suspect by many insurers) as late as the 1970s and 1980s (see Cummins et al. (1983) for life insurance; see Casey et al. (1976); Massachusetts Division of Insurance (1978); and New Jersey Department of Insurance (1981) for auto insurance). When actuaries did finally adopt multivariate statistical techniques, they were used not to score discrete variables (or risk factors) that could then be tallied to arrive at an individualized price (as in credit scoring), but simply to produce more reliable estimates of cell-means calculated by observing the average loss experience of members of a given risk class (e.g., Hsiao et al., 1990). In other words, modern multivariate statistics were made to conform to the requirements of class-based pricing rather than disrupting this system (see Barry, 2020).
We have simplified auto insurance pricing here for ease of presentation. In addition to classifying individuals by driver characteristics, insurers also classify vehicles by the territory in which they are garaged. Each of these two sets of classifications produces a “relativity” – indicating how much greater or lower the propensity to file claims is for drivers with a particular set of characteristics or a car garaged in a particular territory compared to the statewide average. Thus, classification by driver characteristic and by territory produces two different sets of prices, and the key controversy roiling the insurance industry beginning in the mid-1970s (as auto insurance rates quickly inflated) was how to combine them to calculate a fair premium (see Casey et al., 1976; Florida Department of Insurance, 1979; Government Accounting Office, 1979; Massachusetts Division of Insurance, 1978; New Jersey Department of Insurance, 1981).
Roi Livne (2021: 921-22) identifies a particularly striking example of class-based decision-making: the practice of using demographic categories such as age, education, gender, race, and ethnicity to ascertain the wishes of dying individuals when they are unable to communicate these wishes directly (because currently incapacitated and having failed to document their preferences when in a condition to do so).
Put differently, the cell is not itself the unit at which mobilization occurs (as Simon (1988) correctly observes) but is constructed from socially legible characteristics that enable mobilization across cells.
“NOW Insurance Project,” October 29, 1982, MC 623, Folder 5, Box 126, National Organization for Women Legal Defense and Education Fund Records, Schlesinger Library, Radcliffe Institute, Harvard University, Cambridge, MA.
This continued to be the case until the passage of the Affordable Care Act in 2010, which prohibited the use of gender in determining the price of health insurance. The European Union passed comprehensive legislation banning the use of gender as a pricing variable across all lines of insurance in 2012 (Mabbett, 2014), making the case we examine here anomalous not only with respect to other institutional domains in the United States, but also with respect to insurance practices in the international context.
NOW also targeted gender discrimination in health, disability, and life insurance (see Krippner, 2021).
“Sex Bias Alleged in NOW Suit,” Philadelphia Inquirer, August 17, 1984.
Interview with Deborah Ellis conducted by Greta Krippner, February 8, 2017, Rutgers, New Jersey.
“Memo to Preparers of PA NOW Insurance Case from Twiss and Pat Butler,” September 14, 1986, MC 666, Folder 4, Box 363, NOW LDEF Records.
“Aetna: Our Case for Sex Discrimination,” MC 496, Folder 26, Box 117, National Organization for Women Records, Schlesinger Library, Radcliffe Institute, Harvard University, Cambridge, MA.
In fact, NOW believed that insurers adopted gender-based pricing for younger drivers precisely so that they could (falsely) claim that all women got a break on their auto insurance premiums, hence concealing the overcharge paid by women drivers over the age of 25 (see Butler et al., 1988). “Complaint: Pennsylvania NOW versus State Farm,” September 23, 1986, MC 666, Folder 4, Box 363, NOW Records; “Plaintiffs’ Brief in Opposition to Defendants’ and Insurance Department’s Motions for Stay of Proceedings, More Specific Pleadings, and Dismissal of Plaintiffs’ Claims,” November 21, 1986, MC 666, Folder 1, Box 132, NOW Records; “Brief for Petitioners Requesting Review of Insurance Commissioners’ Opinion and Order,” September 14, 1987, MC 496, Folder 25, Box 117, NOW Records.
“Memo to NOW National Board from Sheri O’Dell Re: Insurance Discrimination Activity,” May 1, 1986, MC 666, Folder 4, Box 363, NOW LDEF Records; “Memo to Preparers of PA NOW Insurance Case from Twiss and Pat Butler,” September 14, 1986, MC 666, Folder 4, Box 363, NOW LDEF Records.
“Plaintiffs’ Hearing Brief,” August 14, 1987, MC 496, Folder 24, Box 117, NOW Records.
Paradoxically, when sex was first introduced as a rating variable in the 1950s, insurers seemed to share this understanding. As one history of automobile insurance rating notes, “[T]he first attempt to recognize a statistical difference between young male and female drivers … was not based on the belief that female drivers were necessarily better drivers than male drivers of a comparable age, but rather that the exposure was less with young female drivers because of their infrequent use of a family car as compared with that of a young male driver” (Zoffer, 1959: 158; emphasis added).
“Plaintiffs’ Hearing Brief,” August 14, 1987, MC 496, Folder 24, Box 117, NOW Records.
“Dialogue on Pennsylvania NOW’s Auto Insurance Sex Discrimination Lawsuit,” September 23, 1988, MC 496, Folder 23, Box 117, NOW Records; emphasis added.
Other insurance companies named as defendants in the lawsuit were Nationwide, Allstate, and Liberty Mutual Insurance. In addition, the Insurance Services Office, the state agency that pools data to create standardized risk classifications for use by smaller insurers, was also included in the complaint.
“Complaint: Pennsylvania NOW versus State Farm,” September 23, 1986, MC 666, Folder 4, Box 363, NOW Records.
“Plaintiffs’ Brief in Opposition to Defendants’ and Insurance Department’s Motions for Stay of Proceedings, More Specific Pleadings, and Dismissal of Plaintiffs’ Claims,” November 21, 1986, MC 666, Folder 1, Box 132, NOW Records.
“Letter to Sally Burns from Deborah Ellis,” November 19, 1987, MC 623, Folder 2, Box 128, NOW Records; “Women’s Groups Split on Unisex Car Insurance Rates,” The Pittsburgh Press, September 23, 1988.
“Dialogue on Pennsylvania NOW’s Auto Insurance Sex Discrimination Lawsuit,” September 23, 1988, MC 496, Folder 23, Box 117, NOW Records; “Some Thoughts on True Equality,” Allentown Morning Call, October 2, 1988, MC 496, Folder 23, Box 117, NOW Records.
“Press Strategy and Analysis, PA NOW Auto Insurance Case,” October 14, 1986, MC 666, Folder 4, Box 363, NOW Records.
“Women’s Groups Split on Unisex Car Insurance Rates,” The Pittsburgh Press, September 23, 1988.
“Perspective on Automobile Insurance Pricing,” Presented by Patrick Butler at the National Conference of State Legislatures Conference on the Crisis in the Insurance Market, Boston, Massachusetts, February 24, 1989, MC 663, Folder 14, Box 24, NOW Records.
PA. N.O.W et al. v. PA. Ins. Dept, 122 PA Commw 283 (1988).
As Barbara Brown and Ann Freedman (1975: 46; emphasis added) observed, “As long as the insurance companies group people on a basis that has some consistent predictive value, the group experience will seem correct, and it will be difficult for those who constitute a subgroup with a different risk to identify themselves as such.”
“Dialogue on Pennsylvania NOW’s Auto Insurance Sex Discrimination Lawsuit,” September 23, 1988, MC 496, Folder 23, Box 117, NOW Records; “Some Thoughts on True Equality,” Allentown Morning Call, October 2, 1988, MC 496, Folder 23, Box 117, NOW Records.
We discuss credit scoring here as it was practiced when first widely adopted in the 1970s and not as the technique subsequently evolved in later decades. The most critical development in this regard involved the shift beginning in the second half of the 1980s from custom scorecards constructed from each user’s own loan files to generic (FICO) scores generated from credit-bureau data. This shift coincided with new uses of credit scores that were no longer applied simply to the decision to extend or deny credit, but also used to determine variable prices for loan products (i.e., risk-based pricing). As Martha Poon (2007: 300) notes, the development of generic scores produced an object that was even more fully decontextualized compared to custom scorecards: “In circulating everywhere, in appearing as the same kind of number, in being perpetually recalculated, consumer credit risk calculation is no longer anchored in particular moments or specific places.” Notably, this process of decontextualization has only increased with the advent of “big data,” amplifying and accelerating the processes we describe here. We consider the implications of our argument given more recent developments in artificial intelligence and machine learning in the conclusion.
We could also refer to this decision-making technology as “variable-based.” We prefer attributes to variables because it is the language that creditors themselves use to describe an item of information about an applicant. More precisely, “attributes” are possible values on the variables scored by creditors (e.g., “homeowner” is an attribute of the variable “type of residence”; “18–25” is an attribute of the variable “age”; and so on) (see Lewis, 1994).
There are two distinct issues at play here. First, as a purely technical matter, because creditors analyze data on variables rather than whole persons, even relatively small samples generate sufficient observations to calculate reliable estimates (provided certain independence assumptions hold). This affords creditors considerable flexibility in constructing scoring models especially when compared to insurers, whose reliance on cross-classified data significantly increases the number of observations needed to generate reliable estimates (see footnote 15). Second, the “risk pool” in credit markets is defined by a common score (see Morris, 1966: 55), with no expectation that the individuals who share risk (i.e., pay the same price for credit) hold characteristics in common other than the achieved score. In insurance, by contrast, the risk pool is constituted by the class, which is assumed to be formed by individuals who hold a series of risk-relevant characteristics in common. The expectation that insurance risk classes are constituted by individuals who represent the “same” risk significantly constrains the selection of variables used to construct classes, as we have already noted. Critically, there is no analogous constraint on creditors. In fact, the cells in the credit scoring table can be collapsed for statistical expediency, regardless of whether the resulting groupings make sense in some larger sociological sense (see Lewis, 1994: 55–58). Thus, creditors are not required to select variables that ensure that individuals are “fairly” grouped in the same way as is true for insurers.
This was often claimed as an advantage of credit scoring over the more subjective forms of credit screening it displaced (see Johnson, 1992), but it applies equally well to a comparison with class-based methods of pricing risk such as used in insurance markets.
The following discussion draws on Krippner (2017).
Race, ethnicity, national origin, religion, and other protected classes were added to the statute in amendments passed in 1976.
Letter to Federal Reserve Board from Cynthia Harrison, June 5, 1979, Folder 18, Box 1, Cynthia Harrison Papers, Schlesinger Library, Radcliffe Institute, Harvard University, Cambridge, MA.
Letter to State Coordinators from Cynthia Harrison, July 25, 1978, Folder 28, Box 45, Cynthia Harrison Papers.
So-called “judgmental” credit screening involved a face-to-face interview in which creditors weighed relevant factors subjectively, often relying on “gut feelings” to produce a decision (Stuart, 2003).
As a result, when an individual’s score was tallied across these various characteristics, she might land in a “group” that contained only one person. While such a result would signal the failure of insurance classification – because group experience could not be statistically validated on the basis on one observation – it was in effect the aspiration of credit scoring systems to place each individual in her own class. Nothing better illustrates the different ontologies of insurance and credit – class versus attribute – than the fact that having too few individuals in a class caused endless handwringing in insurance (e.g., Casey et al., 1976) but was a prime objective of credit scoring (Morris, 1966: 54).
Put differently, even if gender were not expressly prohibited as a pricing variable, the way scoring models refract an individual’s group membership across multiple categories such that one is not placed in a single, overarching class with others would make collective mobilization more difficult in the credit scoring context compared to insurance pricing.
“Proposed Rulemaking, Federal Reserve System, Equal Credit Opportunity: Application to Credit Scoring,” Folder 18, Box 1, Cynthia Harrison Papers.
Federal Reserve Board Hearings on Proposed Regulations to Implement the Equal Credit Opportunity Act, Statement of National Organization for Women, July 14, 1975, Folder 22, Box 1, Cynthia Harrison Papers.
“Proposed Rulemaking, Federal Reserve System, Equal Credit Opportunity: Application to Credit Scoring,” Folder 18, Box 1, Cynthia Harrison Papers.
Letter to Federal Reserve Board of Governors from Cynthia Harrison, August 21, 1979, Folder 18, Box 1, Cynthia Harrison Papers.
Letter to Federal Reserve Board of Governors from Cynthia Harrison, August 21, 1979, Folder 18, Box 1, Cynthia Harrison Papers; Letter to Federal Reserve Board of Governors from Edith Canty, June 15, 1979, Folder 19, Box 1, Cynthia Harrison Papers.
Letter to Federal Reserve Board of Governors from Edith Canty, June 15, 1979, Folder 19, Box 1, Cynthia Harrison Papers.
Letter to Linda J. Wilt from M. McLay (New Accounts Department, J.C. Penney), June 20, 1979, Folder 2, Box 1, Cynthia Harrison Papers.
Letter to Federal Reserve Board of Governors from Cynthia Harrison, August 21, 1979, Folder 18, Box 1, Cynthia Harrison Papers.
Letter to Federal Reserve Board of Governors from Cynthia Harrison, June 5, 1979, Folder 18, Box 1, Cynthia Harrison Papers; Letter to Federal Reserve Board of Governors from Edith Canty, June 15, 1979, Folder 19, Box 1, Cynthia Harrison Papers.
Letter to Federal Reserve Board of Governors from Cynthia Harrison, June 5, 1979, Folder 18, Box 1, Cynthia Harrison Papers; emphasis added.
NOW’s Credit Task was abruptly disbanded late in 1979 (see Krippner, 2017 for a fuller history of the Task Force). Note to Cynthia Harrison from Barbara Duke, January 20, 1980, Folder 1, Box 1, Cynthia Harrison Papers.
Interview with Cynthia E. Harrison conducted by Greta Krippner, August 24, 2010, Washington D.C.
This paradox is not entirely novel to the digital era, as Joshua Gamson (1995) long ago noted that anti-discrimination movements undermined the conditions necessary for their continued existence by attempting to eliminate from use the very same categories that were also the source of collective identification and political mobilization. What is arguably new in the digital era is the way algorithmic technologies directly weaken processes of group formation, arguably making Gamson’s paradox all the more acute.
More accurately, the future of the politics of the classification likely lies with artificial intelligence and machine learning – technologies that we expect to accelerate and amplify the trends we associate here with the early development of credit scoring. We elaborate briefly on this point in our discussion below.
Rebecca Elliott’s (2021) work on the Federal Emergency Management Agency’s (FEMA) flood insurance program provides an instructive example here. As Elliot discusses, FEMA’s restructuring of its risk classification scheme created a group constituted by a new category of person, the “flood zone homeowner.” This was an “artificial” group to be sure, and yet when Hurricane Sandy visited destruction on those joined together by FEMA’s new classification, they found a shared identity and a common political purpose expressed in mobilization to “Stop FEMA Now!” As Elliott (2021: 121) pointedly notes, “[T]he flood insurance rate map did not displace or defuse contestation; it instead helped to activate and organize it.”
Notably, class-based technologies may also involve the suppression of salient social identities such as race or gender. But as we observed in our analysis, the operation of forming classes itself tends to allow these identities to “show through” even in circumstances when they have been formally eliminated from classification schemes, thus maintaining potentials for collective identification.
By “machine learning” we reference a range of computational techniques iteratively applied to data to identify patterns in order to make predictions and inform decisions (Hurley & Adebayo, 2016: 160–61).
These techniques, it should be noted, are still experimental in credit markets, where the vast majority of credit decisions continue to rely on FICO scores calculated using more conventional statistical techniques (Hurley & Adebayo, 2016: 155). Jenna Burrell (2016: 11) reports that the company that produces the FICO score has for now avoided adopting machine learning techniques in part because of the difficulty of interpreting the resulting score. Thus, if these techniques represent the future of scoring technologies, this future has yet to fully arrive in credit markets, and appears even more remote in insurance given fundamental tensions between class-based ontologies and “individualized” pricing methods (see especially Barry & Charpentier, 2020).
See discussion of NOW’s credit campaign above.
We are indebted to Roi Livne for the discussion in this paragraph.
Given the social salience of the events that produced the subprime mortgage crisis, the geographic concentration of resulting foreclosures, and the clear discriminatory nature of lending practices (connected to a longer history of such discrimination), we would consider the identity of the subprime borrower especially propitious for political mobilization. In fact, it is difficult to imagine conditions that would be more likely to produce mobilization by a category of persons joined together by a common score, leaving us doubtful that other such mobilizations will materialize in less propitious circumstances.
An important counterpoint here is the way other kinds of algorithms than those we consider here may encourage identity-based political mobilization. We have in mind here recommendation systems that curate content on social media platforms, contributing to the resurgence of identity-based politics in its most virulent form (see Bail, 2021). We are intrigued by the relationship between algorithms that dampen and amplify group identities as a basis for political mobilization, but fully exploring this relationship is beyond the scope of the current article.
This, of course, is not the only way of imagining the relationship between individual and collective in sociology. For an alternative to the conventional view that is particularly amenable to the digital age, see Bruno Latour’s (2002; 2010) generative exploration of the social theory of Gabriel Tarde.
In insisting here on narrative identity as a constitutive feature of personhood, we are drawing on the work of Margaret Somers (1994).
References
Abraham, K. S. (1986). Distributing risk: Insurance, legal theory, and public policy. Yale University Press.
Aitken, R. (2017). ‘All data is credit data’: Constituting the unbanked. Competition and Change, 21(4), 274–300.
Alborn, T. L. (2009). Regulated lives: Life insurance and British society, 1800–1914. University of Toronto Press.
Amoore, L. (2020). Cloud ethics: Algorithms and the attributes of ourselves and others. Duke University Press.
Appel, H. (2015). You are not a loan: Strike debt and the emerging debtors movement. Tikkun, 30(1), 28–30.
Arendt, H. (1998 [1958]). The human condition. Second edition, with an introduction by M. Canovan. University of Chicago Press.
Austin, R. (1983). The insurance classification controversy. University of Pennsylvania Law Review, 131(3), 517–583.
Avraham, R., Logue, K. D., & Schwarcz, D. (2014). Understanding insurance anti-discrimination laws. Southern California Law Review, 87(2), 195–270.
Bail, C. (2021). Breaking the social media prism: How to make our platforms less polarizing. Princeton University Press.
Baker, T. (2002). Risk, insurance, and the social construction of responsibility. In T. Baker & J. Simon (Eds.), Embracing risk: The changing culture of risk and responsibility (pp. 33–51). University of Chicago Press.
Barry, L. (2019). Rationality of digital governmentality. Journal for Cultural Research, 23(4), 365–380.
Barry, L. (2020). Insurance, big data, and changing conceptions of fairness. European Journal of Sociology, 61(2), 159–184.
Barry, L., & Charpentier, A. (2020). Personalization as a promise: Can big data change the practice of insurance? Big Data & Society, 7(1), 1–12.
Barry, L., & Fisher, E. (2019). Digital audiences and the deconstruction of the collective. Subjectivity, 12, 210–227.
Beer, D. (2017). The social power of algorithms. Information, Communication, and Society, 20(1), 1–13.
Benjamin, R. (2019). Race after technology: Abolitionist tools for the new Jim code. Polity.
Bennett, W. L., & Segerberg, A. (2012). The logic of connective action: Digital media and the personalization of contentious politics. Information, Communication, and Society, 15(5), 739–768.
Bouk, D. (2015). How our days became numbered: Risk and the rise of the statistical individual. University of Chicago Press.
Bourdieu, P. (1984). Distinction: A social critique of the judgement of taste. Harvard University Press.
Bourdieu, P. (1985). The social space and the genesis of groups. Theory and Society, 14(6), 723–744.
Brandel, R. E. (1976). New dangers arise in point scoring, but you can't afford to be without it. Banking, (March), 86–93.
Brayne, S. (2020). Predict and surveil: Data, discretion, and the future of policing. Oxford University Press.
Brayne, S., & Christin, A. (2021). Technologies of crime prediction: The reception of algorithms in policing and criminal courts. Social Problems, 68(3), 608–624.
Brilmayer, L., Hekeler, R. W., Laycock, D., & Sullivan, T. A. (1980). Sex discrimination in employer-sponsored insurance plans: A legal and demographic analysis. The University of Chicago Law Review, 47(3), 505–560.
Brown, B., & Freedman, A. E. (1975). Sex averaging and the Equal Rights Amendment. Women’s Rights Law Reporter, 2, 35–49.
Brubaker, R. (2005). Ethnicity without groups. In J. Adams, E. S. Clemens, & S. Orloff (Eds.), Remaking modernity (pp. 470–492). Duke University Press.
Brubaker, R. (2015). Difference and inequality. In Grounds for difference (pp. 10–47). Harvard University Press.
Burrell, J. (2016). How the machine ‘thinks’: Understanding opacity in machine learning. Big Data and Society, 3(1), 1–12.
Butler, J. (2000). Appearances aside. California Law Review, 88(1), 55–63.
Butler, P., Butler, T., & Williams, L. L. (1988). Sex-divided mileage, accident, and insurance cost data show that auto insurers overcharge most women. Journal of Insurance Regulation, 6(3), 244–284.
Casey, B., Pezier, J., & Spetzler, C. (1976). The role of risk classification in property and casualty insurance: A study of the risk assessment process. Stanford Research Institute.
Cevolini, A., & Esposito, E. (2020). From pool to profile: Social consequences of algorithmic prediction in insurance. Big Data & Society, 7(2), 1–11.
Chang, L., & Fairley, W. B. (1978). An estimation model for multivariate insurance rate classification. In Automobile insurance risk classification: Equity & Accuracy (pp. 25–55). Massachusetts Division of Insurance.
Cheney-Lippold, J. (2017). We are data: Algorithms and the making of our digital selves. NYU Press.
Christin, A. (2020). What can data do: A typology of mechanisms. International Journal of Communication, 14, 1115–1134.
Clark, G. W. (1999). Betting on lives: The culture of life insurance in England, 1695–1775. Manchester University Press.
Couldry, N. (2015). The myth of ‘us’: Digital networks, political change, and the production of collectivity. Information, Communication, and Society, 18(6), 608–626.
Cummins, J. D., Smith, B. D., Vance, R. N., & VanDerhei, J. L. (1983). Risk classification in life insurance. Huebner international series on risk, insurance, and economic security. Kluwer-Nijhoff.
Daston, L. (2022). Rules: A short history of what we live by. Princeton University Press.
Demkovich, L. (1977). New equal credit regulations are creating headaches. National Journal, (March 5), 354–359.
Desrosières, A. (1998). The politics of large numbers: A history of statistical reasoning. Harvard University Press.
Dourish, P. (2016). Algorithms and their others: Algorithmic culture in context. Big Data & Society, 3(2), 1–11.
Durand, D. (1941). Risk elements in consumer instalment financing. National Bureau of Economic Research.
Durkheim, E. (1995 [1912]). The elementary forms of religious life. Free Press.
Durkheim, E. (2014 [1893]). The division of labor in society. Free Press.
Elliott, R. (2021). Underwater: Loss, flood insurance, and the moral economy of climate change in the United States. Columbia University Press.
Espeland, W. N., & Sauder, M. (2016). Engines of anxiety: Academic rankings, reputation, and accountability. Russell Sage Foundation.
Ettlinger, N. (2018). Algorithmic affordances for productive resistance. Big Data & Society, 5(1), 1–13.
Eubanks, V. (2018). Automating inequality: How high-tech tools profile, police, and punish the poor. St. Martin’s Press.
Ewald, F. (2020). The birth of solidarity: The history of the French welfare state. Duke University Press.
Feeley, M. M., & Simon, J. (1992). The new penology: Note on the emerging strategy of corrections and its implications. Criminology, 30(4), 449–474.
Ferriera, J., Shapiro, A. F., & VanSlyke, O. E. (1978). Classification models. Record of Society of Actuaries, 4(1), 115–132.
Florida Department of Insurance. (1979). An evaluation of risk classification systems in automobile insurance (report prepared by D.J. Nye, H.M. Cushinsky, D.R. Forrester, J.M. Kulik, and P. Tenenbaum).
Fourcade, M. (2016). Ordinalization. Sociological Theory, 34(3), 175–195.
Fourcade, M. (2021). Ordinal citizenship. British Journal of Sociology, 72(2), 154–173.
Fourcade, M., & Healy, K. (2013a). Classification situations: Life chances in the neoliberal era. Accounting, Organizations, and Society, 38(8), 559–572.
Fourcade, M., & Healy, K. (2013b). Credit scores and the moralization of inequality. Paper presented at the Annual Meeting of the American Sociological Association.
Fourcade, M., & Johns, F. (2020). Loops, ladders, and links: The recursivity of social and machine learning. Theory and Society, 49(5–6), 803–832.
Fry, H. (2018). Hello world: Being human in the age of algorithms. W.W. Norton & Company.
Gamson, J. (1995). Must identity movements self-destruct? A queer dilemma. Social Problems, 42(3), 390–407.
Gerber, J. S. (1975). The economic and actuarial aspects of selection and classification. Forum, 10(4), 1205–1230.
Gillespie, T. (2014). The relevance of algorithms. In T. Gillespie, P. J. Boczkowski, & K. A. Foot (Eds.), Media technologies: Essays on communication, materiality, and society (pp. 167–193). MIT Press.
Goldberg, C. (2007). Citizens and paupers: Relief, rights, and race from the Freedmen’s bureau to workfare. Chicago University Press.
Gosden, P. H. J. H. (1961). The friendly societies in England, 1815–1875. Manchester University Press.
Government Accounting Office. (1979). Issues and needed improvements in state regulation of the insurance business: Report to the Congress.
Gowri, A. (1997). The irony of insurance: community and commodity. Ph.D. Thesis, University of Southern California.
Graham, H. D. (1990). The civil rights era: Origins and development of national policy, 1960–1972. Oxford University Press.
Hacking, I. (1986). Making up people. In T. Heller, M. Sosna, & D. Wellberry (Eds.), Reconstructing individualism: Autonomy, individuality, and the self in Western thought (pp. 222–236). Stanford University Press.
Hacking, I. (1990). The taming of chance. Cambridge University Press.
Hand, D. J., & Henley, W. E. (1997). Statistical classification methods in consumer credit scoring: A review. Journal of the Royal Statistical Society, 160 (Part 3), 523–541.
Harcourt, B. (2015). Exposed: Desire and disobedience in the digital age. Harvard University Press.
Heen, M. (2014). Nondiscrimination in insurance: The next chapter. Georgia Law Review, 49(1), 1–77.
Heimer, C. (1985). Reactive risk and rational action: Managing moral hazard in insurance contracts. University of California Press.
Hirschman, D., & Bosk, E. A. (2020). Standardizing biases: Selection devices and the quantification of race. Sociology of Race and Ethnicity, 6(3), 348–364.
Hopkins, E. (1995). Working-class self-help in nineteenth-century England: Responses to industrialization. St. Martin’s Press.
Horan, C. (2021). Insurance era: Risk, governance, and the privatization of security in postwar America. University of Chicago Press.
Hsia, D. C. (1978). Credit scoring and the Equal Credit Opportunity Act. Hastings Law Journal, 30(2), 371–448.
Hsiao, C., Kim, C., & Taylor, G. (1990). A statistical perspective on insurance rate-making. Journal of Econometrics, 44(1–2), 5–24.
Hunter, N. (2008). Risk governance and deliberative democracy in health care. Georgetown Public Law and Legal Theory Research Paper, No. 10–75.
Hurley, M., & Adebayo, J. (2016). Credit scoring in the era of big data. Yale Journal of Law and Technology, 18, 148–216.
Hyman, L. (2011). Debtor nation: A history of America in red ink. Princeton University Press.
Igo, S. E. (2007). The averaged American: Surveys, citizens, and the making of a mass public. Harvard University Press.
Isin, E., & Ruppert, E. (2020). Being digital citizens (second edition). Rowman and Littlefield.
Johnson, R. W. (1992). Legal, social, and economic issues in implementing scoring in the U.S. In C. Thomas, J. N. Crook, & D. B. Edelman (Eds.), Credit scoring and credit control (pp. 19–32). Oxford University Press.
Josephson, H. (1960). Discrimination: A study of recent developments in American life insurance. Wesley Press.
Kiviat, B. (2019). The moral limits of predictive practices: The case of credit-based insurance scores. American Sociological Review, 84(6), 1134–1158.
König, P. D., & Wenzelburger, G. (2021). When politicization stops algorithms in criminal justice. The British Journal of Criminology, 61(3), 832–851.
Krippner, G. R. (2017). Democracy of credit: Ownership and the politics of credit access in late twentieth-century America. American Journal of Sociology, 123(1), 1–47.
Krippner, G. R. (2021). Gendered market devices: The persistence of gender discrimination in insurance markets. Unpublished manuscript, Sociology Department. University of Michigan.
Lamont, M., & Molnar, V. (2002). The study of boundaries in the social sciences. Annual Review of Sociology, 28, 167–195.
Lamont, M., Beljean, S., & Clair, M. (2014). What is missing? Cultural processes and causal pathways to inequality. Socio-Economic Review, 12(3), 573–608.
Latour, B. (2002). Gabriel Tarde and the end of the social. In P. Joyce (Ed.), The social in question: New bearings in history and the social sciences (pp. 117–132). Routledge.
Latour, B. (2010). Tarde’s idea of quantification. In M. Candea (Ed.), The social after Gabriel Tarde: Debates and assessments (pp. 145–162). Routledge.
Lauer, J. (2017). Creditworthy: A history of consumer surveillance and financial identity in America. Columbia University Press.
Lautzenheiser, B. (1976). Sex and the single table: Equal monthly retirement income for the sexes? Employee Benefits Journal, 2(1), 8–13, 38–44.
Lehtonen, T. K., & Liukko, J. (2011). The forms and limits of insurance solidarity. Journal of Business Ethics, 103, 33–44.
Lengwiler, M. (2009). Double standards: The history of standardizing humans in modern life insurance. In M. Lampland & S. L. Star (Eds.), Standards and their stories: How quantifying, classifying, and formalizing practices shape everyday life (pp. 95–113). Cornell University Press.
Levy, J. (2012). Freaks of fortune: The emerging world of capitalism and risk in America. Harvard University Press.
Lewis, E. M. (1994). An introduction to credit scoring. Athena Press.
Livne, R. (2021). Toward a sociology of finitude: Life, death, and the question of limits. Theory and Society, 50(6), 891–934.
Mabbett, D. (2014). Polanyi in Brussels or Luxembourg? Social rights and market regulation in European insurance. Regulation & Governance, 8(2), 186–202.
MacCormick, J. (2012). Nine algorithms that changed the future: The ingenious ideas that drive today’s computers. Princeton University Press.
Marron, D. (2007). ‘Lending by numbers’: Credit scoring and the constitution of risk within American consumer credit. Economy and Society, 36(1), 103–133.
Marron, D. (2009). Consumer credit in the United States: A sociological perspective from the 19th century to the present. Palgrave Macmillan.
Massachusetts Division of Insurance. (1978). Automobile insurance risk classification: Equity & Accuracy.
Massey, D. S. (2007). Categorically unequal: The American stratification system. Russell Sage Foundation.
Massey, D. S., Rugh, J. S., Steil, J. P., & Albright, L. (2016). Riding the stagecoach to hell: A qualitative analysis of racial discrimination in mortgage lending. City & Community, 15(2), 118–136.
Mauss, M. (1985). A category of the human mind: The notion of person; the notion of self. In M. Carrithers, S. Collins, & S. Lukes (Eds.), The category of the person: Anthropology, philosophy, and history (pp. 1–25). Cambridge University Press.
McFall, L. (2019). Personalizing solidarity? The role of self-tracking in health insurance pricing. Economy and Society, 48(1), 52–76.
McFall, L., & Moor, L. (2018). Who, or what, is insurtech personalizing?: Persons, prices and the historical classifications of risk. Distinktion. Journal of Social Theory, 19(2), 193–213.
Milan, S. (2015). When algorithms shape collective action: Social media and the dynamics of cloud protesting. Social Media and Society, 1(2), 1–10.
Monk, E. (2022). Inequality without groups: Contemporary theories of categories, intersectional typicality, and the disaggregation of difference. Sociological Theory, 40(1), 3–27.
Moor, L., & Lury, C. (2018). Price and the person: Markets, discrimination, and personhood. Journal of Cultural Economy, 11(6), 501–513.
Mora, C. G. (2014). Making Hispanics: How activists, bureaucrats, and media constructed a new American. University of Chicago Press.
Morris, R. A. (1966). Credit analysis: An O.R. approach. Management Services, 3 (March–April), 52–57.
Myers, J. H., & Forgy, E. W. (1963). The development of numerical evaluation systems. American Statistical Association Journal, (September), 799–806.
New Jersey Department of Insurance. (1981). Hearing on automobile insurance classifications and related methodologies, Final determination, analysis and report.
Noble, S. U. (2018). Algorithms of oppression: How search engines reinforce racism. New York University Press.
O’Malley, P. (1996). Risk and responsibility. In A. Barry, T. Osborne, & N. Rose (Eds.), Foucault and political reason: Liberalism, neoliberalism, and rationalities of government (pp. 189–207). Taylor and Francis.
O’Neil, C. (2016). Weapons of math destruction: How big data increases inequality and threatens democracy. Crown Press.
Pasquale, F. (2015). The black box society: The secret algorithms that control money and information. Harvard University Press.
Poon, M. (2007). Scorecards as devices for consumer credit: The case of Fair, Isaac & Company Incorporated. The Sociological Review, 55, 284–306.
Poon, M. (2012). What lenders see: A pre-history of the FICO® credit scoring system. Ph.D. Thesis, University of California, San Diego.
Porter, T. (1986). The rise of statistical thinking, 1820–1900. Princeton University Press.
Reid, C. K. (2017). Financialization and the subprime subject: The experiences of homeowners during California’s housing boom. Housing Studies, 32(6), 793–815.
Ridgeway, C. L. (2011). Framed by gender: How gender inequality persists in the modern world. Oxford University Press.
Ridgeway, C. L. (2019). Status: Why is it everywhere? Why does it matter? Russell Sage Foundation.
Rieder, B. (2016). Big data and the paradox of diversity. Digital Culture and Society, 2(2), 39–54.
Rieder, B. (2017). Scrutinizing an algorithmic technique: The Bayes classifier as interested reading of reality. Information, Communication & Society, 20(1), 100–117.
Rieder, B. (2020). Engines of order: A mechanology of algorithmic techniques. University of Amsterdam.
Robinson + Yu (2014). Knowing the score: New data, underwriting, and marketing in the consumer credit marketplace: A guide for financial inclusion stakeholders. https://www.upturn.org/work/knowing-the-score-new-data-underwriting-and-marketing-in-the-consumer-credit/
Rodgers, D. T. (2011). Age of fracture. Belknap Press.
Rodríguez-Muñiz, M. (2021). Figures of the future: Latino civil rights and the politics of demographic change. Princeton University Press.
Rona-Tas, A. (2017). The off-label use of consumer credit ratings. Historical Social Research, 42(1), 52–76.
Root, E. K. (1915). The decreasing mortality from tuberculosis. In proceedings of the Association of Life Insurance Medical Directors of America from the twenty-third to and including the twenty-fifth annual meeting (pp. 358–374). The Knickerbocker Press.
Rosanvallon, P. (2000). The new social question: Rethinking the welfare state. Princeton University Press.
Rothstein, W. (2003). Public health and the risk factor. University of Rochester Press.
Rouvroy, A. (2013). The end(s) of critique: Data behaviourism versus due process. In M. Hildebrandt & K. de Vries (Eds.), Privacy, due process, and the computational turn: The philosophy of law meets the philosophy of technology (pp. 143–168). Routledge.
Rouvroy, A., & Berns, T. (2013). Translated by E. Libbrecht. Algorithmic governmentality and the prospects of emancipation. Reseaux, 177(1), 163–196.
Ruppert, E. (2012). The governmental topologies of database devices. Theory, Culture, and Society, 29(4–5), 116–136.
Sandvig, C., Hamilton, K., Karahalios, K., & Langbort, C. (2014). Auditing algorithms: Research methods for detecting discrimination on internet platforms. Data and Discrimination: Converting Critical Concerns into Productive Inquiry, 22, 4349–4357.
Sartre, J. P. (1976). Critique of dialectical reason (Eds.) J. Ree; translated by A. Sheridan Smith. New Left Books.
Seaver, N. (2017). Algorithms as culture: Some tactics for the ethnography of algorithmic systems. Big Data & Society, 4(2), 1–12.
Seaver, N. (2018). What should an anthropology of algorithms do? Cultural Anthropology, 33(3), 375–385.
Shilton, E. (2012). Insuring inequality: Sex-based mortality tables and women’s retirement income. Queen’s Law Journal, 37(2), 383–435.
Simon, J. (1987). The emergence of a risk society: Insurance, law, and the state. Socialist Review, 95 (Fall/Winter), 60–89.
Simon, J. (1988). The ideological effects of actuarial practices. Law & Society Review, 22(4), 771–800.
Somers, M. R. (1994). The narrative constitution of identity: A relational and network approach. Theory and Society, 23(5), 605–650.
Stalson, J. O. (1942). Marketing life insurance; Its history in America. Harvard University Press.
Starr, S. B. (2014). Evidence-based sentencing and the scientific rationalization of discrimination. Stanford Law Review, 66(4), 803–872.
Stone, J. (1978). Excerpt from the opinion, findings and decision on 1978 automobile insurance rates. In Automobile insurance risk classification: Equity & Accuracy (pp. 144–205). Massachusetts Division of Insurance.
Stone, D. (1993). The struggle for the soul of health insurance. Journal of Health Politics, Policy and Law, 18(2), 287–317.
Stone, D. (2002). Beyond moral hazard: Insurance as moral opportunity. In T. Baker & J. Simon (Eds.), Embracing risk: The changing culture of insurance and responsibility (pp. 52–79). University of Chicago Press.
Stuart, G. (2003). Discriminating risk: The U.S. mortgage lending industry in the twentieth century. Cornell University Press.
Sunstein, C. (2017). #Republic: Divided democracy in the age of social media. Princeton University Press.
Taylor, W. F. (1980). Meeting the Equal Credit Opportunity Act’s specificity requirement: Judgmental and statistical scoring systems. Buffalo Law Review, 29 (Winter), 73–130.
Taylor, K. Y. (2019). Race for profit: How banks and the real estate industry undermined black homeownership. University of North Carolina Press.
Thomas, W. (2013). The mathematical management of uncertainty in Britain and America in the Second World War and after. In A. Durand, L. Mazliak, & R. Tazziol (Eds.), Des mathématiciens et des guerres: Histoires de confrontations (XIXe-XXe Siècle) (pp. 71–79). CNRS Éditions.
Thomas, W. (2015). Rational action: The sciences of policy in Britain and America, 1940–1960. MIT Press.
Thurston, C. (2018). At the boundaries of homeownership: Credit, discrimination, and the American state. Cambridge University Press.
Tilly, C. (1998). Durable inequality. University of California Press.
Trumbull, G. (2014). Consumer lending in France and America: Credit and welfare. Cambridge University Press.
Tufekci, Z. (2014). Engineering the public: Big data, surveillance and computational politics. First Monday, 19(7).
Tufekci, Z. (2017). Twitter and tear gas: The power and fragility of networked protest. Yale University Press.
Weisberg, H., & Tomberlin, T. (1982). A statistical perspective on actuarial methods for estimating pure premiums from cross-classified data. Journal of Risk and Insurance, 49(4), 539–563.
Wells, H. L. (1963). Numerical pointing plans for evaluating consumer credit risks. In Second consumer credit symposium (pp. 4–21). University of Pennsylvania.
Wherry, F. F., K. S. Seefeldt, & A. S. Alvarez. (2019). Credit where it’s due: Rethinking financial citizenship. Russell Sage Foundation.
Witt, J. F. (2004). The accidental republic: Crippled workingmen, destitute widows, and the remaking of American law. Harvard University Press.
Wortham, L. (1985). Insurance classification: Too important to be left to the actuaries. University of Michigan Journal of Law Reform, 19, 349–423.
Young, I. M. (1994). Gender as seriality: Thinking about women as serial collective. Signs, 19(3), 713–738.
Zelizer, V. A. R. (1979). Morals and markets: The development of life insurance in the United States. Columbia University Press.
ZestFinance. (n.d.). Five building blocks for compliant AI-driven lending. Accessed June 7, 2022: https://www.zest.ai/resources/five-building-blocks-for-compliant-ai-driven-lending.
Zoffer, J. (1959). The history of automobile liability insurance rating. University of Pittsburgh Press.
Zuboff, S. (2019). The age of surveillance capitalism: The fight for a human future at the new frontier of power. Public Affairs.
Acknowledgments
We gratefully acknowledge the contributions of Barbara Kiviat, Marion Fourcade, Angelina Grigoryeva, Kieran Healy, Sasha Killewald, Sandra Levitsky, Roi Livne, Jeff Lockhart, Davon Norris, Sarah Quinn, Kelly Russell, Nathan Wilmers, Andreas Wimmer, Nathan Worrell, and the participants in the Social Theory and Economic Sociology workshops in the Sociology Department at the University of Michigan. We also thank audiences at MaxPo in Paris, the Sociology Department at Washington University, the Sociology Department at the University of Toronto, and the Harvard-MIT Economic Sociology Workshop. Finally, we benefited from the opportunity to present versions of this paper at the annual meetings of the Society for the Advancement of Socio-Economics (2016) and the American Sociological Association (2017, 2018). This research has been generously supported by the Interdisciplinary Committee on Organizational Studies, the Institute for Research on Women and Gender, and the College of Literature, Sciences, and the Arts, all at the University of Michigan.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Krippner, G.R., Hirschman, D. The person of the category: the pricing of risk and the politics of classification in insurance and credit. Theor Soc 51, 685–727 (2022). https://doi.org/10.1007/s11186-022-09500-5
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11186-022-09500-5