Ethics of Behavioral Intervention: Autonomy, Consent, and Behavioral Power

Last Updated June 22, 2026

Ethics of Behavioral Intervention appears here not as an afterthought to behavioral science, but as one of its central responsibilities. Whenever an institution designs a default, removes or adds friction, changes a message, uses a social norm, personalizes a platform, introduces a reward system, or tests a behavioral intervention, it is making a claim about influence, welfare, responsibility, and power.

The field matters because behavioral science is often attractive precisely when people are vulnerable to context. People are busy, uncertain, tired, stressed, overloaded, financially constrained, socially influenced, digitally tracked, or institutionally dependent. Behavioral ethics asks whether interventions respect those conditions or exploit them.

GitHub Repository

The companion repository for this knowledge series should be created as ethics-behavioral-intervention-code under the Content-Catalyst-LLC GitHub organization. It should use article-level folders under articles/, shared reusable methods under _shared/, and documented synthetic-data workflows so the pillar can support conceptual articles, ethical review checklists, intervention-risk matrices, dark-pattern audits, consent-flow analysis, equity diagnostics, algorithmic nudging examples, public-accountability templates, and reproducible demonstrations of responsible behavioral governance.

Behavioral Intervention as Ethical Practice

Behavioral intervention is ethical practice because it changes the conditions under which people act. An intervention may make one option easier, another harder, one message more salient, one path more likely, one behavior more visible, one default automatic, or one form more difficult to exit. These choices are never purely technical. They express judgments about welfare, agency, responsibility, and institutional power.

A behavioral intervention may be designed to help people act on their own goals. A reminder may reduce missed appointments. A simplified form may improve access. A default may protect people from costly inaction. A social norm message may support cooperation. In these cases, behavioral design can expand practical agency by reducing unnecessary burden.

But the same tools can also be used to manipulate. Friction can be added to cancellation. Defaults can hide costly commitments. Rankings can exploit status anxiety. Variable rewards can capture attention. Personalized messages can target vulnerability. Behavioral intervention is therefore not good or bad by category. Its ethical status depends on purpose, transparency, context, consequences, power, consent, and accountability.

Influence, Power, and Human Agency

Ethics of Behavioral Intervention begins with a realistic view of influence. Human behavior is always shaped by environments, norms, institutions, incentives, design, and social expectations. There is no perfectly neutral choice environment. But the absence of neutrality does not mean every influence is legitimate. It means influence must be examined and governed.

Power matters because not all actors influence from the same position. A government agency, employer, school, hospital, landlord, bank, platform, insurer, or public authority can shape behavior under conditions of dependency. The person being influenced may have limited exit, limited information, limited time, or limited bargaining power. Ethical analysis must therefore attend to asymmetry.

Agency is not protected merely by leaving a formal choice available. A user may technically be able to opt out while the process is confusing, hidden, costly, or designed to discourage exit. A citizen may technically be able to access a service while administrative burden makes access unrealistic. A worker may technically be free to refuse while workplace pressure makes refusal costly. Behavioral ethics asks whether agency is practically meaningful, not merely formally preserved.

Behavioral Ethics and Institutional Accountability

Behavioral ethics is institutional because interventions are often designed, approved, deployed, measured, and revised by organizations. Ethical responsibility cannot rest only with individual designers or researchers. It belongs to the system that authorizes behavioral influence.

Responsible institutions should be able to answer basic questions. What behavior is being targeted? Why? Who benefits? Who may be harmed? What evidence supports the intervention? What alternatives were considered? Is the intervention transparent? Can people refuse, contest, or exit? Are vulnerable groups protected? Are effects distributed fairly? Is there an audit trail? Who is accountable if the intervention causes harm?

Institutional accountability also requires documentation. Behavioral interventions should not disappear into informal design decisions, hidden optimization systems, or undocumented experiments. When behavioral science shapes public life, workplaces, platforms, or services, its assumptions and effects should be reviewable.

What Ethics of Behavioral Intervention Studies

Ethics of Behavioral Intervention studies the moral and governance questions raised by behaviorally informed systems. At the individual level, it examines autonomy, consent, welfare, dignity, attention, vulnerability, manipulation, coercion, and informed choice. At the social level, it studies stigma, social pressure, norm enforcement, inequity, group targeting, public trust, and collective consequences.

At the institutional level, the field studies behavioral public policy, workplace influence, educational interventions, health interventions, platform design, administrative burden, defaults, service access, regulatory design, dark patterns, and accountability. At the technological level, it studies algorithmic nudging, personalization, recommender systems, predictive targeting, attention capture, behavioral analytics, and automated experimentation.

At the methodological level, it studies consent, deception, research ethics, experimentation, privacy, data governance, evaluation, unintended consequences, transparency, reproducibility, and intervention review. Its central concern is how behavioral science can be used without undermining the people it claims to help.

What This Pillar Covers

This pillar covers the major domains through which ethics of behavioral intervention is studied and applied. It includes autonomy, consent, transparency, paternalism, manipulation, coercion, welfare, dignity, public justification, democratic legitimacy, equity, vulnerability, distributional effects, stigma, dark patterns, digital choice architecture, algorithmic nudging, platform influence, behavioral public policy, administrative burden, workplace behavior systems, health behavior interventions, education interventions, sustainability interventions, research ethics, privacy, experimental design, governance, accountability, auditability, and institutional review.

These domains differ in setting, but they share a common problem: behavioral influence changes the practical conditions of agency. A default in a public benefit system, a reward loop in an app, a compliance message from a tax agency, a ranking system in a workplace, and a social norm campaign in public health all raise ethical questions about influence, power, and justification.

The series also treats ethics as constructive rather than merely restrictive. Ethical review should not only say no. It should help design better interventions: clearer, fairer, more transparent, more accountable, more agency-supporting, and more sensitive to context.

Mathematics, Computation, and Modeling in Behavioral Ethics

Mathematics and computation can support behavioral ethics by making assumptions, tradeoffs, distributional effects, and risks explicit. They cannot decide what is ethical by themselves, but they can help institutions examine consequences before and after intervention.

A simple distributional treatment-effect model can represent intervention effects by group:

A disparity measure can compare the strongest and weakest group-level effects:

An intervention-risk score can combine potential harm, opacity, vulnerability, and power asymmetry:

where \(H\) represents potential harm, \(O\) opacity, \(V\) vulnerability, \(P\) power asymmetry, \(C\) consent or contestability, and \(w\) values are review weights that should be publicly justified rather than hidden.

A practical agency-support measure can compare ease of entry and ease of exit:

These formulas are not moral calculators. They are scaffolds for review. They help ask whether an intervention has unequal effects, whether harm is concentrated, whether people understand the design, whether exit is meaningful, and whether institutional power is being used responsibly.

R supports intervention evaluation, distributional analysis, fairness diagnostics, survey analysis, and impact assessment. Python supports audit pipelines, simulation, digital interface analysis, dark-pattern detection scaffolds, and behavioral risk modeling. SQL supports intervention records, consent states, user-flow logs, audit trails, complaint records, subgroup outcomes, and governance documentation. Other languages can support typed review schemas, formal constraints, reproducible infrastructure, and accountable decision systems.

Major Domains of Ethics of Behavioral Intervention

Ethics of Behavioral Intervention includes several major domains. Autonomy and consent examine whether people understand and can meaningfully accept, refuse, or exit influence. Manipulation and coercion examine when influence bypasses reflection, exploits vulnerability, or pressures behavior through asymmetry. Paternalism examines when institutions steer people for claimed welfare benefits and what justification is required.

Equity and vulnerability examine whether interventions burden or target groups unfairly. Transparency and accountability examine whether behavioral systems can be explained, audited, contested, and governed. Digital ethics examines dark patterns, algorithmic nudging, personalization, engagement optimization, recommender systems, and platform design. Public-policy ethics examines legitimacy, democratic oversight, public trust, and the responsible use of behavioral science by governments.

Research ethics examines consent, deception, experimentation, privacy, human subjects protection, and responsible evaluation. Together, these domains make behavioral ethics a field of both philosophical analysis and institutional practice.

Why Ethics of Behavioral Intervention Matters

Ethics of Behavioral Intervention matters because behavioral tools are powerful precisely where people are limited. Defaults matter because people are busy or uncertain. Friction matters because people have limited time and energy. Salience matters because attention is scarce. Social norms matter because belonging is important. Personalization matters because people differ. These same realities make behavioral influence useful and ethically dangerous.

The field also matters because behavioral interventions often scale. A small design change in a platform, public agency, financial product, or health system can affect thousands or millions of people. Effects may be invisible to users. Harms may be distributed unevenly. Benefits may be claimed by institutions while costs fall on individuals.

Most importantly, ethics matters because behavioral science should not merely make systems more effective at steering people. It should help build systems that support agency, dignity, fairness, trust, and public value. Behavioral effectiveness without ethical accountability is not enough.

Behavioral Ethics, Agency, and Public Trust

Public trust is central to behavioral ethics. People are more likely to accept behavioral interventions when they believe institutions are transparent, legitimate, fair, competent, and acting in the public interest. Hidden or manipulative influence can erode trust even if it produces short-term behavioral effects.

Agency is also central. Ethical behavioral design should help people understand options, act on their own goals, avoid unnecessary burden, and retain meaningful exit. It should not rely on confusion, exhaustion, shame, addiction, fear, or hidden friction. A behaviorally informed system should make good choices easier without making dissent, refusal, or exit unreasonable.

This is especially important in public systems, workplaces, schools, healthcare, and digital platforms, where people may depend on institutions. Ethical behavioral intervention should be judged not only by what people do, but by what kind of relationship the intervention creates between people and institutions.

Ethics of Behavioral Intervention Pillar Map

The map below organizes the Ethics of Behavioral Intervention knowledge series into conceptual domains, moving from foundations of influence and autonomy toward consent, paternalism, manipulation, equity, digital systems, public policy, research ethics, governance, climate policy, AI governance, digital rights, organizational power, public knowledge, and future directions. The article titles below are intentionally left unlinked until their individual pages exist.

Measurement, Review, and Responsible Governance

Ethical behavioral intervention requires measurement, but not only measurement of behavioral success. An intervention should be evaluated for benefits, harms, comprehension, autonomy, trust, equity, distributional effects, unintended consequences, and practical exit. If an intervention changes behavior while reducing dignity, obscuring choice, or increasing unequal burden, its success is ethically incomplete.

Responsible review should happen before and after deployment. Before deployment, institutions should identify the behavior being targeted, affected populations, power relationships, consent conditions, possible harms, alternatives, transparency requirements, and accountability structures. After deployment, they should measure actual effects, subgroup differences, complaints, misuse, long-term outcomes, and trust.

Governance should also include documentation. Behavioral systems should preserve records of intervention design, evidence, ethical review, deployment conditions, changes over time, evaluation results, and responsible owners. Behavioral influence should not become invisible simply because it is embedded in design.

Behavioral Ethics, Technology, and Digital Systems

Digital systems have made behavioral intervention more continuous, personalized, scalable, and opaque. Platforms can test designs rapidly, personalize messages, change defaults, rank content, trigger notifications, recommend actions, and optimize flows based on behavioral data. This makes digital behavioral ethics one of the most important areas of contemporary governance.

Digital choice environments can support users. They can simplify complex tasks, provide timely feedback, increase accessibility, and help people act on goals they value. But digital systems can also manipulate attention, obscure consent, personalize pressure, create addictive loops, bury exit options, and optimize institutional metrics over human welfare.

Ethical digital behavioral design should therefore ask what the system optimizes, what users understand, what they can refuse, how vulnerable groups are protected, how effects are audited, and who is accountable. Engagement is not the same as wellbeing. Retention is not the same as value. Conversion is not the same as consent.

Behavioral Ethics, Sustainability, and Public Systems

Sustainability interventions raise distinctive ethical questions. Behavioral tools can support conservation, energy efficiency, public transit use, waste reduction, climate adaptation, and collective action. But they can also shift responsibility onto individuals while leaving structural barriers intact. A nudge toward sustainable behavior is ethically weak if the sustainable option is unavailable, unaffordable, unsafe, or socially unsupported.

Behavioral ethics in sustainability should therefore distinguish support from blame. Good interventions reduce barriers, make sustainable action easier, provide trustworthy feedback, support collective participation, and align with public investment and structural change. Poor interventions use behavioral language to avoid deeper institutional responsibility.

Public systems also raise questions of legitimacy. Governments may use behavioral science to support public goals, but public power requires public justification. Sustainability behavior should be encouraged through transparent, fair, participatory, and accountable systems, not covert steering or symbolic burden shifting.

R Section: Simulating Distributional Effects and Intervention Risk

For analytical readers, R is useful for evaluating distributional effects, subgroup outcomes, and intervention risk. The example below simulates a behavioral intervention that improves average outcomes but has unequal effects across groups. It is not real behavioral data. It is a reproducible scaffold for thinking clearly about why ethical evaluation must look beyond average effects.

# Synthetic ethics of behavioral intervention example in R
# Educational example only.
# This script evaluates average and group-specific intervention effects.

# install.packages(c("tidyverse", "broom", "scales"))

library(tidyverse)
library(broom)
library(scales)

set.seed(10701)

n_people <- 3600

ethics_data <- tibble(
  person_id = 1:n_people,
  group = sample(
    c("group_a", "group_b", "group_c"),
    size = n_people,
    replace = TRUE,
    prob = c(0.45, 0.35, 0.20)
  ),
  baseline_access = runif(n_people, 0.10, 0.95),
  comprehension = runif(n_people, 0.15, 0.95),
  institutional_trust = runif(n_people, 0.10, 0.95),
  vulnerability_index = runif(n_people, 0.00, 1.00),
  treatment = rbinom(n_people, 1, 0.50)
) |>
  mutate(
    group_treatment_effect = case_when(
      group == "group_a" ~ 0.55,
      group == "group_b" ~ 0.25,
      group == "group_c" ~ -0.10
    ),

    outcome_score =
      -1.80 +
      0.95 * baseline_access +
      0.70 * comprehension +
      0.65 * institutional_trust -
      0.45 * vulnerability_index +
      treatment * group_treatment_effect,

    outcome_probability = 1 / (1 + exp(-outcome_score)),
    beneficial_outcome = rbinom(n_people, 1, outcome_probability),

    opacity = runif(n_people, 0.10, 0.90),
    power_asymmetry = runif(n_people, 0.10, 0.95),
    meaningful_consent = runif(n_people, 0.05, 0.90),

    ethical_risk_score =
      0.30 * vulnerability_index +
      0.25 * opacity +
      0.30 * power_asymmetry -
      0.25 * meaningful_consent
  )

average_effect <- ethics_data |>
  group_by(treatment) |>
  summarise(
    outcome_rate = mean(beneficial_outcome),
    mean_risk = mean(ethical_risk_score),
    .groups = "drop"
  )

print(average_effect)

group_effects <- ethics_data |>
  group_by(group, treatment) |>
  summarise(
    outcome_rate = mean(beneficial_outcome),
    mean_risk = mean(ethical_risk_score),
    .groups = "drop"
  ) |>
  pivot_wider(
    names_from = treatment,
    values_from = c(outcome_rate, mean_risk),
    names_prefix = "treatment_"
  ) |>
  mutate(
    estimated_effect = outcome_rate_treatment_1 - outcome_rate_treatment_0,
    risk_difference = mean_risk_treatment_1 - mean_risk_treatment_0
  )

print(group_effects)

ethics_model <- glm(
  beneficial_outcome ~ treatment * group +
    baseline_access + comprehension +
    institutional_trust + vulnerability_index,
  data = ethics_data,
  family = binomial(link = "logit")
)

print(tidy(ethics_model, conf.int = TRUE, exponentiate = TRUE))

ggplot(group_effects, aes(x = group, y = estimated_effect)) +
  geom_col() +
  labs(
    title = "Synthetic Group-Specific Intervention Effects",
    x = "Group",
    y = "Estimated effect on beneficial outcome"
  ) +
  scale_y_continuous(labels = percent_format()) +
  theme_minimal()

ggplot(ethics_data, aes(x = ethical_risk_score, y = outcome_probability)) +
  geom_point(alpha = 0.18) +
  geom_smooth(method = "loess", se = FALSE) +
  labs(
    title = "Synthetic Relationship Between Ethical Risk and Outcome Probability",
    x = "Ethical risk score",
    y = "Predicted beneficial outcome probability"
  ) +
  scale_y_continuous(labels = percent_format()) +
  theme_minimal()

This workflow models a core ethical intuition: an intervention can look positive on average while producing unequal, weak, or harmful effects for specific groups. Ethical evaluation requires subgroup analysis, harm measurement, and attention to power, consent, vulnerability, and transparency.

Python Section: Modeling Ethical Review of Behavioral Interventions

Python is useful for building intervention-review workflows and risk-screening tools. The example below creates a synthetic ethical review table for behavioral interventions. The goal is not to automate ethics, but to demonstrate how institutions can document key review dimensions before deployment.

# Synthetic ethics of behavioral intervention review in Python
# Educational example only.
# This script scores intervention risk dimensions for review and documentation.

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

rng = np.random.default_rng(10701)

interventions = [
    "default_enrollment",
    "reminder_message",
    "social_norm_message",
    "personalized_prompt",
    "friction_reduction",
    "ranking_system",
    "gamified_streak",
    "cancellation_flow",
    "risk_warning",
    "adaptive_recommendation"
]

n = len(interventions)

review = pd.DataFrame({
    "intervention": interventions,
    "expected_public_benefit": rng.uniform(0.20, 0.95, n),
    "potential_harm": rng.uniform(0.05, 0.85, n),
    "opacity": rng.uniform(0.05, 0.90, n),
    "power_asymmetry": rng.uniform(0.10, 0.95, n),
    "vulnerability_exposure": rng.uniform(0.05, 0.90, n),
    "meaningful_consent": rng.uniform(0.05, 0.95, n),
    "exit_availability": rng.uniform(0.05, 0.95, n),
    "equity_review_quality": rng.uniform(0.05, 0.95, n),
    "auditability": rng.uniform(0.05, 0.95, n)
})

review["risk_score"] = (
    0.22 * review["potential_harm"]
    + 0.18 * review["opacity"]
    + 0.20 * review["power_asymmetry"]
    + 0.20 * review["vulnerability_exposure"]
    - 0.12 * review["meaningful_consent"]
    - 0.04 * review["exit_availability"]
    - 0.02 * review["equity_review_quality"]
    - 0.02 * review["auditability"]
)

review["agency_support_score"] = (
    0.30 * review["meaningful_consent"]
    + 0.25 * review["exit_availability"]
    + 0.20 * review["auditability"]
    + 0.15 * review["equity_review_quality"]
    + 0.10 * (1 - review["opacity"])
)

review["review_category"] = pd.cut(
    review["risk_score"],
    bins=[-1, 0.20, 0.40, 1],
    labels=["lower_review_priority", "moderate_review_priority", "high_review_priority"]
)

print(review.sort_values("risk_score", ascending=False))

plt.figure(figsize=(10, 6))
plt.bar(review["intervention"], review["risk_score"])
plt.xticks(rotation=35, ha="right")
plt.ylabel("Synthetic risk score")
plt.title("Synthetic Ethical Risk Score by Intervention")
plt.tight_layout()
plt.show()

plt.figure(figsize=(10, 6))
plt.scatter(review["risk_score"], review["agency_support_score"])
for _, row in review.iterrows():
    plt.annotate(
        row["intervention"],
        (row["risk_score"], row["agency_support_score"]),
        fontsize=8,
        alpha=0.80
    )

plt.xlabel("Synthetic ethical risk score")
plt.ylabel("Synthetic agency support score")
plt.title("Risk and Agency Support in Behavioral Intervention Review")
plt.tight_layout()
plt.show()

review.to_csv("ethical_behavioral_intervention_review.csv", index=False)

For designers, researchers, policymakers, and institutional leaders, the key lesson is that ethics should be documented, not improvised. A review workflow should not replace human judgment, but it can make assumptions visible and create a record of responsibility.

Interpretive Limits and Ethical Cautions

Ethical analysis cannot be reduced to checklists or scores. A high benefit score does not automatically justify a manipulative intervention. A low-risk classification does not remove the need for context. A transparent intervention can still be coercive. A formally optional choice can still be practically constrained. Ethical judgment requires interpretation, deliberation, and accountability.

Behavioral ethics must also avoid treating people as abstract users. People encounter interventions as citizens, patients, students, workers, parents, consumers, community members, and rights-bearing persons. Context changes what an intervention means. A reminder from a friend is different from a reminder from a government agency. A default in a charity sign-up is different from a default in a high-stakes public service. A personalized prompt in education is different from a personalized prompt in political manipulation.

The field is strongest when it treats people not as behavioral targets, but as participants in systems that should be understandable, contestable, fair, and humane. Behavioral science should support agency, not quietly replace it.

Ethics of Behavioral Intervention in a Wider Intellectual Context

Ethics of Behavioral Intervention belongs to a long intellectual history of inquiry into autonomy, persuasion, paternalism, consent, manipulation, power, welfare, law, public reason, and institutional legitimacy. Philosophers have debated freedom, coercion, dignity, and the ethics of influence. Legal scholars have examined consent, consumer protection, administrative power, and rights. Political theorists have studied democratic legitimacy and public justification. Psychologists and behavioral economists have studied bounded rationality, defaults, social influence, and decision environments.

Modern behavioral intervention brings these traditions together in practical systems. Governments use nudges. Platforms optimize attention. Employers design incentives. Schools shape motivation. Health systems design adherence pathways. Sustainability programs encourage behavior change. Each setting raises ethical questions about how behavioral science should be used and governed.

For Sustainable Catalyst, this field is essential because knowledge, design, and policy should not merely become more effective. They should become more accountable. Ethics of Behavioral Intervention provides one of the clearest bridges between behavioral science, public trust, technology governance, institutional design, sustainability, and human dignity.

Related Reading

• Behavioral Science & Behavioral Psychology
• Choice Architecture and Nudging
• Behavioral Public Policy
• Behavioral Research Methods
• Social Norms and Behavioral Influence
• Behavior Change and Habit Formation
• Motivation, Reinforcement, and Learning
• Behavioral Economics
• Moral Psychology
• Ethics
• Institutions & Governance
• Technology & Systems Intelligence
• Public Policy
• Sustainable Development

Further Reading

• Bovens, L. (2009) ‘The ethics of nudge’, in Grüne-Yanoff, T. and Hansson, S.O. (eds.) Preference Change: Approaches from Philosophy, Economics and Psychology. Dordrecht: Springer, pp. 207–219.
• Hansen, P.G. and Jespersen, A.M. (2013) ‘Nudge and the manipulation of choice: A framework for the responsible use of the nudge approach to behaviour change in public policy’, European Journal of Risk Regulation, 4(1), pp. 3–28.
• Hausman, D.M. and Welch, B. (2010) ‘Debate: To nudge or not to nudge’, Journal of Political Philosophy, 18(1), pp. 123–136.
• Rebonato, R. (2012) Taking Liberties: A Critical Examination of Libertarian Paternalism. Basingstoke: Palgrave Macmillan.
• Sunstein, C.R. (2014) Why Nudge? The Politics of Libertarian Paternalism. New Haven, CT: Yale University Press.
• Sunstein, C.R. (2016) The Ethics of Influence: Government in the Age of Behavioral Science. Cambridge: Cambridge University Press.
• Thaler, R.H. and Sunstein, C.R. (2021) Nudge: The Final Edition. New Haven, CT: Yale University Press.
• Yeung, K. (2017) ‘Hypernudge: Big Data as a mode of regulation by design’, Information, Communication & Society, 20(1), pp. 118–136.
• Zuboff, S. (2019) The Age of Surveillance Capitalism. New York: PublicAffairs.

References

• Bovens, L. (2009) ‘The ethics of nudge’, in Grüne-Yanoff, T. and Hansson, S.O. (eds.) Preference Change: Approaches from Philosophy, Economics and Psychology. Dordrecht: Springer, pp. 207–219.
• Hansen, P.G. and Jespersen, A.M. (2013) ‘Nudge and the manipulation of choice: A framework for the responsible use of the nudge approach to behaviour change in public policy’, European Journal of Risk Regulation, 4(1), pp. 3–28. Available at: https://doi.org/10.1017/S1867299X00002762 (Accessed: 21 June 2026).
• Hausman, D.M. and Welch, B. (2010) ‘Debate: To nudge or not to nudge’, Journal of Political Philosophy, 18(1), pp. 123–136. Available at: https://doi.org/10.1111/j.1467-9760.2009.00351.x (Accessed: 21 June 2026).
• Mills, S. (2015) ‘The ethics of nudging’, Behavioural Public Policy, 1(1), pp. 1–28.
• Rebonato, R. (2012) Taking Liberties: A Critical Examination of Libertarian Paternalism. Basingstoke: Palgrave Macmillan.
• Sunstein, C.R. (2014) Why Nudge? The Politics of Libertarian Paternalism. New Haven, CT: Yale University Press.
• Sunstein, C.R. (2016) The Ethics of Influence: Government in the Age of Behavioral Science. Cambridge: Cambridge University Press.
• Thaler, R.H. and Sunstein, C.R. (2021) Nudge: The Final Edition. New Haven, CT: Yale University Press.
• Yeung, K. (2017) ‘Hypernudge: Big Data as a mode of regulation by design’, Information, Communication & Society, 20(1), pp. 118–136. Available at: https://doi.org/10.1080/1369118X.2016.1186713 (Accessed: 21 June 2026).
• Zuboff, S. (2019) The Age of Surveillance Capitalism. New York: PublicAffairs.

Back to top ↑