Ethics, Power, and Inclusion in Design Thinking

Last Updated May 28, 2026

Ethics, power, and inclusion are not secondary concerns in design thinking. They determine whether design thinking becomes a serious human-centered practice or a polished method for reproducing the assumptions of powerful institutions. A design process may use empathy interviews, journey maps, workshops, prototypes, and testing, yet still fail ethically if it ignores who defines the problem, who has decision power, who carries risk, who is excluded, and who benefits from the final design.

Design thinking is often introduced as a process for understanding people, reframing problems, generating ideas, prototyping solutions, and learning through iteration. That process can be valuable. But design does not happen in a vacuum. It happens inside organizations, markets, public agencies, schools, hospitals, platforms, communities, infrastructures, and political systems. Every design decision distributes attention, access, burden, dignity, risk, time, cost, visibility, control, and opportunity.

This means design thinking must be examined through ethics and power. Human-centered design is not automatically just because it begins with users. Inclusion is not guaranteed because a team invites people into a workshop. Empathy is not enough if the people most affected by a design have no authority over the outcome. Testing is not enough if the measures of success ignore harm, exclusion, surveillance, exploitation, or institutional distrust.

A more serious design-thinking practice asks harder questions. Who is centered? Who is missing? Who is asked to share vulnerability without receiving power? Who benefits from the design? Who bears the cost of failure? What histories of exclusion, racism, ableism, colonialism, gendered power, class inequality, or institutional harm shape the design context? What forms of knowledge are treated as credible? What trade-offs are hidden behind language like innovation, efficiency, engagement, scale, or transformation?

Ethical design thinking does not abandon creativity. It deepens it. It expands design from problem solving to responsibility: responsibility for process, participation, evidence, implementation, governance, and consequence. It treats inclusion not as representation alone, but as a redesign of decision power, access, accountability, and care.

Ethics, power, and inclusion connect directly to what design thinking is, human-centered problem solving, empathy and stakeholder research, contextual inquiry and synthesis, problem framing, insight generation, prototyping, testing and validation, iteration and experimentation, service design, behavioral design, strategy, co-design and participatory design, public policy, and organizational innovation.

What Ethics, Power, and Inclusion Mean in Design Thinking

Ethics in design thinking concerns the responsibilities designers and organizations have toward the people affected by their work. It asks whether the design process respects dignity, autonomy, safety, privacy, accessibility, fairness, cultural context, and public consequence. It also asks whether the design outcome reduces harm or merely makes an institution more efficient at achieving its own goals.

Power concerns who has the authority to define problems, choose methods, interpret evidence, allocate resources, make trade-offs, decide what counts as success, and control implementation. Power is present in every design process, even when it is not named. A design team may ask users for stories but reserve all authority for executives. A public agency may invite community input but ignore the conclusions. A platform may test user behavior while hiding how the data will be used. A healthcare service may claim patient-centeredness while forcing patients to navigate inaccessible systems.

Inclusion concerns who is invited, who is heard, who is believed, who is accommodated, who is compensated, who can influence decisions, and who benefits from the result. Inclusion is not only about demographic variety in a workshop. It is about access, language, disability, time, trust, safety, decision power, cultural knowledge, historical context, and consequences.

Ethics, power, and inclusion make design thinking more rigorous. They prevent design from becoming a technique for making flawed systems more attractive. They also protect design teams from mistaking participation for justice, empathy for accountability, or innovation for improvement.

Back to top ↑

Why Ethics Matters for Design Thinking

Ethics matters because design thinking often works on real systems that affect real lives: public services, healthcare, education, finance, housing, employment, digital platforms, workplace systems, transportation, civic participation, AI tools, and institutional decision processes. These systems can help people, but they can also exclude, manipulate, surveil, burden, stigmatize, misclassify, or disempower them.

Design thinking becomes ethically dangerous when it treats human experience as input for organizational goals without examining the goals themselves. A team may improve a loan application flow without questioning predatory lending. It may make a surveillance tool easier to use without questioning surveillance. It may optimize public-benefit compliance without reducing administrative burden. It may increase engagement with a platform that exploits attention. It may redesign policing, border control, or workplace monitoring without confronting harm.

Ethics also matters because design thinking is often used under the language of empathy, innovation, and human-centeredness. These words can create moral cover. A process can appear caring while still serving narrow institutional interests. Ethical design thinking asks not only whether people were consulted, but whether consultation changed anything. It asks not only whether the solution is desirable, but desirable for whom and at whose expense.

The central ethical question is not “Did we use design thinking correctly?” It is “Did the design process and outcome respect the people and communities affected by it?”

Back to top ↑

Power in the Design Process

Power enters design thinking long before a prototype appears. It enters when a sponsor defines the brief, when a team chooses who counts as a user, when research questions are written, when participants are recruited, when findings are interpreted, when some stories become insights and others become noise, when budgets are allocated, when trade-offs are made, and when implementation decisions are finalized.

Power is not always malicious. It can be structural, procedural, professional, technical, economic, or cultural. A design team may have power because it controls the methods. Executives may have power because they control resources. Engineers may have power because they control technical feasibility. Policy teams may have power because they define eligibility. Data teams may have power because they define measurement. Users may be called central while having little authority over any of these decisions.

Design thinking becomes more ethical when it treats power as a design material. Teams should map power just as carefully as they map journeys. They should ask where authority sits, whose knowledge is discounted, what constraints are negotiable, what decisions have already been made, and whether participation can meaningfully influence the outcome.

If the answer is no, the process should be honest about its limits. A consultation process should not be called co-design if participants cannot shape decisions. A workshop should not be called inclusive if it excludes people who cannot attend, speak safely, or influence outcomes. Naming power is the beginning of ethical design practice.

Back to top ↑

Inclusion Beyond Representation

Inclusion is often reduced to representation: making sure different kinds of people are present in the design process. Representation matters, but it is not enough. A process can include diverse participants and still be inaccessible, extractive, unsafe, tokenistic, rushed, uncompensated, or powerless. Inclusion must be evaluated by influence, not presence alone.

Inclusive design thinking asks what conditions people need in order to participate meaningfully. This includes language access, disability access, scheduling, childcare, transportation, compensation, digital access, cultural safety, trauma awareness, privacy, trust, facilitation quality, and clear explanation of how input will be used. It also includes the right not to participate, especially when participation asks people to relive harm without any guarantee of change.

Inclusion also requires attention to intersectionality. People do not experience systems through a single identity or constraint. Disability, race, language, class, gender, age, immigration status, geography, caregiving, employment, digital access, and institutional trust can interact. A design that appears inclusive on one dimension may still exclude people at the intersection of several burdens.

The practical test is simple: did inclusion change the design, the decision process, the implementation plan, or the accountability structure? If not, inclusion may have been symbolic rather than substantive.

Back to top ↑

The Limits of Empathy

Empathy is central to design thinking, but empathy has limits. It can help designers listen, notice, and care. It can challenge assumptions. It can reveal lived experience. But empathy can also become performative, extractive, or superficial if it is treated as a substitute for power-sharing, structural analysis, or accountability.

One risk is empathy without consequence. A design team may listen to painful stories, convert them into insights, and then deliver a solution that changes little. Another risk is empathy as emotional consumption: participants share vulnerability while professionals gain understanding, confidence, or innovation credentials. A third risk is empathy as overconfidence: designers may believe they understand another person’s experience after a brief interview or workshop.

Empathy should therefore be paired with responsibility. The goal is not to feel what another person feels; that is often impossible. The goal is to understand enough to act with care, humility, and accountability. Ethical design thinking treats empathy as the beginning of obligation, not the completion of it.

A stronger practice might be called accountable empathy: listening that is consent-based, compensated, contextual, participatory, and tied to visible change.

Back to top ↑

Ethical Problem Framing

Problem framing is one of the most ethically significant steps in design thinking. The frame determines what counts as the problem, who is responsible, which solutions are imaginable, and what evidence is considered relevant. A harmful frame can make injustice appear as inconvenience, exclusion appear as low engagement, distrust appear as communication failure, and structural barriers appear as individual behavior problems.

Ethical problem framing asks whose problem is being solved and whether the frame shifts responsibility away from the institution. For example, “users fail to complete the form” frames the problem as user behavior. “The form and eligibility process impose excessive cognitive and documentation burden” frames the problem as service design and policy. “Employees resist change” frames employees as obstacles. “The change increases workload without participation, support, or trust” frames the organization as responsible for implementation conditions.

Ethical problem framing requires dissent. Teams should invite alternative frames, especially from people most affected by the system. They should ask what the sponsor’s frame hides, what historical context matters, what harms are normalized, and what would change if the problem were framed from the perspective of the most burdened group.

The quality of a design process depends heavily on the quality of its frame. A human-centered method cannot correct a dehumanizing frame unless the frame itself is open to challenge.

Back to top ↑

Participation, Co-Design, and Decision Power

Participation can strengthen design thinking, but only when it is honest about power. There is a difference between consultation, participation, co-design, co-production, and community control. These terms are often used loosely, but they represent different levels of influence.

A project may consult people by asking for feedback. It may involve people by inviting them to workshops. It may co-design by giving participants meaningful influence over problem framing, idea generation, prioritization, and testing. It may co-produce by involving affected people in delivery and governance. It may support community control when affected communities set the agenda and hold decision power.

Ethical design thinking does not pretend that every project can or should use the same level of participation. Some constraints may be real. But the level of participation should be named honestly. Calling a consultation “co-design” inflates legitimacy and misrepresents power.

Participation also requires resources. People need time, compensation, accessible formats, language support, preparation, facilitation, feedback, and clarity about decisions. Without these, participation can shift labor onto communities while institutions retain control.

Back to top ↑

Accessibility, Disability, and Inclusive Design

Accessibility is not a final compliance checklist. It is a design foundation. A design-thinking process that does not include disability access, assistive technologies, plain language, multiple channels, flexible formats, sensory needs, cognitive accessibility, and real testing with disabled people is incomplete.

Inclusive design also recognizes that disability is not only located in individual bodies. Barriers are produced by environments, systems, interfaces, policies, assumptions, and social arrangements. A person may be disabled by a form that cannot be read by a screen reader, a meeting that lacks captions, a service that requires phone calls, a policy that demands in-person visits, or an app that assumes steady attention and fine motor control.

Accessibility strengthens design for everyone, but that should not erase disabled people as the source of expertise. Ethical inclusive design centers the knowledge of people who experience barriers most directly. It also avoids using disability as inspiration while excluding disabled people from authorship, employment, compensation, decision-making, or ownership.

In design thinking, accessibility must appear at the start of research design, not after a solution is nearly finished. It belongs in recruitment, consent, facilitation, prototyping, testing, service delivery, measurement, and governance.

Back to top ↑

Design Justice and Marginalized Knowledge

Design justice asks design to move beyond inclusion toward liberation, accountability, and community-led practice. It challenges the assumption that designers should be the central authors of solutions for communities affected by injustice. Instead, it asks how design can support communities in sustaining, healing, resisting harm, and building the worlds they need.

This matters because design has often served powerful systems: colonial administration, industrial extraction, exclusionary urban planning, racialized surveillance, ableist infrastructure, gendered products, exploitative platforms, financial predation, and institutional control. Design thinking cannot be ethical if it ignores the histories and systems that design has helped build.

Design justice does not mean that professional design skills are irrelevant. It means those skills must be repositioned. Designers can facilitate, visualize, prototype, synthesize, test, document, and build. But they should not assume that method gives them moral authority. In contexts of unequal power, design expertise should be used in service of affected communities, not as a substitute for their leadership.

For design thinking, this is a major shift. The question changes from “How might we solve this problem for users?” to “Who should define this problem, who should lead the work, and how can design resources support accountable change?”

Back to top ↑

Ethics in Service Design and Public Policy

Service design and public policy make ethics concrete. A public service can be legally available but practically inaccessible. A benefit can exist on paper but require documentation, digital access, language fluency, time, confidence, and persistence that many people do not have. A healthcare pathway can be clinically sound but emotionally and administratively punishing. A school system can offer support while placing the burden of navigation on families already under stress.

Design thinking can improve these systems if it examines burden, access, dignity, trust, and accountability. But it can also be misused if it merely makes burdensome systems more efficient. Ethical public-service design asks whether the service should be redesigned at the level of policy, eligibility, documentation, staffing, communication, data, and appeal rights—not only interface or messaging.

Ethics in public design must include public value. A service should not only be usable; it should be legitimate, accountable, accessible, fair, and repairable. People should understand decisions, challenge errors, receive support, and see how feedback changes the system.

Design thinking can support better public institutions when it is tied to rights, governance, and public accountability. Without that, it risks becoming an attractive method for smoothing over deeper institutional failure.

Back to top ↑

Data, AI, Surveillance, and Behavioral Power

Data and AI intensify the ethical stakes of design thinking. Designers now shape systems that classify people, recommend actions, personalize experiences, predict behavior, automate decisions, monitor workers, target messages, optimize engagement, and generate content. These systems can improve access and learning, but they can also reproduce bias, obscure accountability, manipulate behavior, and expand surveillance.

Design thinking must therefore ask whether AI or data-driven design should be used at all, not only how it should be made usable. A harmful system can be user-friendly. A surveillance tool can have a beautiful interface. A biased model can be embedded in a smooth workflow. A manipulative platform can test well on engagement metrics.

AI-assisted design research also needs caution. AI can summarize interviews, cluster themes, generate personas, simulate users, or suggest design ideas. But it can flatten minority experiences, hallucinate patterns, erase context, and make synthetic user representations appear more authoritative than real participants. Designers should not replace affected people with AI-generated approximations of affected people.

Ethical design thinking treats data and AI as governance problems, not only design tools. It asks who controls the model, who can contest outputs, how errors are corrected, what harms are monitored, and whether affected people have meaningful power over the system.

Back to top ↑

Organizational Ethics and Design Governance

Ethical design thinking requires organizational support. Individual designers may care deeply about inclusion and justice, but they often work inside institutions that reward speed, scale, growth, efficiency, persuasion, risk avoidance, or executive preference. If governance does not support ethical practice, ethical design becomes fragile.

Design governance defines how decisions are made, reviewed, documented, challenged, and corrected. It determines whether ethical concerns have authority or remain advisory. It also determines whether research findings can change strategy, whether communities can influence implementation, and whether harms are monitored after launch.

Governance turns ethical language into operational practice. Without governance, ethics depends on individual courage. With governance, ethical review becomes part of the system: documented, funded, timed, authorized, and connected to consequences.

Design thinking should therefore include governance design. The question is not only “What should we design?” It is also “What institutional arrangements will keep this design accountable over time?”

Back to top ↑

Measuring Inclusion, Burden, and Harm

Ethical design thinking needs measurement, but not the narrow measurement of conversion, satisfaction, or engagement alone. A design may increase completion while increasing anxiety. It may reduce internal cost while shifting burden to users. It may improve average satisfaction while excluding high-burden groups. It may raise engagement while harming autonomy or attention. It may appear efficient while reducing trust.

Measurement should include inclusion, burden, dignity, accessibility, equity, trust, error recovery, and harm. It should also be disaggregated. Average outcomes can hide severe failures for disabled users, low-income users, limited-English users, older adults, rural users, frontline staff, caregivers, or people with low institutional trust.

Measurement should not be only extractive. Communities and participants should help define what success means. Frontline staff should help identify hidden labor. People harmed by existing systems should help identify what repair would look like. Ethical measurement includes the right metrics and the right governance around those metrics.

The test is not whether a dashboard looks rigorous. The test is whether evidence can change decisions.

Back to top ↑

Failure, Repair, and Accountability

No design process can eliminate failure. Ethical design thinking therefore needs repair. When a design harms people, excludes them, misclassifies them, burdens them, or fails to deliver what was promised, the organization should have mechanisms for acknowledgement, correction, compensation, redesign, and accountability.

Repair is different from iteration. Iteration asks how to improve the design. Repair asks what responsibility the organization has to people harmed by the design. This is especially important in public services, healthcare, education, finance, AI systems, workplace systems, and other high-stakes contexts.

Accountability should be designed before harm occurs. People should know how to ask questions, correct errors, appeal decisions, report harm, and receive a response. Design teams should know who owns remediation. Leaders should know what evidence triggers stopping, pausing, or redesigning a system.

Ethical design thinking is not proven by avoiding all failure. It is proven by how failure is anticipated, monitored, acknowledged, and repaired.

Back to top ↑

Critiques and Limits

Ethical design thinking has limits. It cannot solve structural injustice by itself. It cannot substitute for law, redistribution, labor rights, disability rights, democratic governance, public accountability, institutional reform, or community power. It can help reveal harms and prototype alternatives, but it cannot overcome political or economic constraints unless organizations are willing to change them.

There is also a risk that ethics becomes another design aesthetic. Teams may add an ethics workshop, inclusion checklist, or justice language without changing decision power. This can make design thinking more legitimate while leaving harmful structures intact.

The limits of ethical design thinking do not make the practice useless. They make humility necessary. Design thinking should not claim to solve every social problem. It should clarify where design can help, where it cannot, and where broader institutional or political change is required.

A serious ethical practice knows when to design, when to stop, when to refuse, when to escalate, and when to support others who should lead.

Back to top ↑

Cross-Pillar Connections

Ethics, power, and inclusion connect design thinking to several broader fields. Stewardship and ethics provide language for responsibility, care, dignity, justice, and long-term consequence. Institutions and governance explain how authority, legitimacy, accountability, and public trust shape design outcomes. Social psychology helps explain group dynamics, bias, identity, belonging, and social influence.

Behavioral design raises questions about autonomy, manipulation, defaults, friction, and choice architecture. Service design reveals how burden, access, recovery, and dignity appear across real journeys. Data systems and AI raise issues of privacy, classification, prediction, bias, and surveillance. Co-design and participatory design provide methods for shifting design authority toward affected people.

The broader lesson is that design thinking is not merely a creativity process. It is a way of intervening in systems that already contain power. Ethical design thinking begins when that power is named, studied, and made accountable.

Back to top ↑

Mathematical Lens: Modeling Inclusion, Burden, and Ethical Risk

Ethics cannot be reduced to equations, but models can make assumptions explicit. A simple inclusion score can combine access, influence, safety, compensation, representation, and accountability.

Burden can be modeled as a combination of time, cognitive effort, emotional stress, documentation, uncertainty, and coordination work.

Ethical risk can be modeled as the expected severity of harm multiplied by exposure and adjusted for detectability, reversibility, and accountability.

The purpose of these models is not to automate ethics. It is to make the design team’s assumptions visible. If an organization cannot say who carries burden, who has influence, who can detect harm, and who can demand repair, then the design process is not ready to claim ethical seriousness.

Back to top ↑

R Workflow: Inclusion, Burden, and Power Analysis

The R workflow below models inclusion, burden, influence, and ethical risk across stakeholder groups. It is designed as a decision-support tool for reflection, not as a substitute for participatory review.

# Install packages if needed.
# install.packages(c("tidyverse", "scales"))

library(tidyverse)
library(scales)

stakeholder_groups <- tibble(
  group = c(
    "Confident digital users",
    "Disabled users",
    "Limited English users",
    "Low digital access users",
    "Frontline staff",
    "Caregivers and intermediaries",
    "Low-trust community members"
  ),
  access = c(8.2, 5.0, 5.4, 4.8, 7.2, 6.4, 5.8),
  voice = c(6.8, 5.2, 5.0, 4.8, 6.0, 5.8, 4.6),
  safety = c(7.6, 5.8, 5.6, 5.4, 6.2, 6.0, 4.8),
  compensation = c(5.0, 4.8, 4.6, 4.4, 5.2, 5.0, 4.2),
  representation = c(7.4, 5.4, 5.2, 5.0, 6.4, 5.8, 4.8),
  accountability = c(6.4, 5.0, 5.0, 4.6, 5.8, 5.4, 4.4),
  time_burden = c(3.2, 7.2, 6.8, 7.0, 6.4, 7.4, 6.8),
  cognitive_burden = c(3.4, 7.4, 7.2, 7.6, 6.6, 6.8, 7.0),
  emotional_burden = c(3.0, 6.8, 6.4, 6.6, 6.8, 7.0, 7.8),
  documentation_burden = c(3.8, 7.0, 6.6, 7.2, 5.8, 6.8, 7.0),
  uncertainty_burden = c(3.6, 7.2, 7.0, 7.4, 6.2, 7.0, 7.8),
  affectedness = c(0.55, 0.92, 0.88, 0.86, 0.72, 0.78, 0.84)
)

group_scores <- stakeholder_groups %>%
  mutate(
    inclusion_score =
      0.20 * access +
      0.20 * voice +
      0.16 * safety +
      0.14 * compensation +
      0.14 * representation +
      0.16 * accountability,
    burden_score =
      0.20 * time_burden +
      0.24 * cognitive_burden +
      0.20 * emotional_burden +
      0.18 * documentation_burden +
      0.18 * uncertainty_burden,
    power_gap =
      10 - ((0.45 * voice) + (0.35 * accountability) + (0.20 * compensation)),
    ethical_attention_priority =
      0.34 * affectedness * burden_score +
      0.28 * affectedness * power_gap +
      0.22 * (10 - inclusion_score) +
      0.16 * (10 - access)
  ) %>%
  arrange(desc(ethical_attention_priority))

print(group_scores)

risk_review <- tibble(
  design_decision = c(
    "Digital-first access",
    "AI-assisted case prioritization",
    "Behavioral reminder campaign",
    "Reduced human support",
    "Community co-design process",
    "Data-driven personalization"
  ),
  harm_severity = c(0.72, 0.86, 0.42, 0.76, 0.28, 0.70),
  probability = c(0.54, 0.48, 0.40, 0.58, 0.30, 0.46),
  exposure = c(0.80, 0.62, 0.74, 0.68, 0.42, 0.66),
  detectability = c(0.46, 0.38, 0.66, 0.42, 0.72, 0.44),
  accountability = c(0.42, 0.36, 0.58, 0.40, 0.76, 0.38)
)

risk_scores <- risk_review %>%
  mutate(
    ethical_risk =
      harm_severity *
      probability *
      exposure *
      (1 - detectability) *
      (1 - accountability),
    review_priority =
      0.36 * ethical_risk +
      0.24 * harm_severity +
      0.18 * exposure +
      0.12 * (1 - accountability) +
      0.10 * (1 - detectability)
  ) %>%
  arrange(desc(review_priority))

print(risk_scores)

ggplot(group_scores, aes(x = reorder(group, ethical_attention_priority), y = ethical_attention_priority)) +
  geom_col() +
  coord_flip() +
  labs(
    title = "Ethical Attention Priority by Stakeholder Group",
    x = "Stakeholder group",
    y = "Ethical attention priority"
  ) +
  theme_minimal(base_size = 12)

write_csv(group_scores, "inclusion_burden_power_scores.csv")
write_csv(risk_scores, "ethical_design_risk_scores.csv")

This workflow helps teams identify where inclusion is weak, where burden is high, where power gaps are greatest, and where design decisions require deeper ethical review before testing or scaling.

Back to top ↑

Python Workflow: Ethical Risk and Inclusion Simulation

The Python workflow below evaluates ethical risk under uncertainty. It estimates which design decisions remain high-risk when assumptions about harm, exposure, detectability, and accountability vary.

# Install packages if needed:
# pip install pandas numpy matplotlib scipy

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

decisions = pd.DataFrame({
    "design_decision": [
        "Digital-first access",
        "AI-assisted case prioritization",
        "Behavioral reminder campaign",
        "Reduced human support",
        "Community co-design process",
        "Data-driven personalization"
    ],
    "harm_severity": [0.72, 0.86, 0.42, 0.76, 0.28, 0.70],
    "probability": [0.54, 0.48, 0.40, 0.58, 0.30, 0.46],
    "exposure": [0.80, 0.62, 0.74, 0.68, 0.42, 0.66],
    "detectability": [0.46, 0.38, 0.66, 0.42, 0.72, 0.44],
    "accountability": [0.42, 0.36, 0.58, 0.40, 0.76, 0.38],
    "inclusion_strength": [0.52, 0.46, 0.62, 0.44, 0.78, 0.48]
})

def score_ethics(df):
    result = df.copy()

    result["ethical_risk"] = (
        result["harm_severity"] *
        result["probability"] *
        result["exposure"] *
        (1 - result["detectability"]) *
        (1 - result["accountability"])
    )

    result["review_priority"] = (
        0.34 * result["ethical_risk"] +
        0.22 * result["harm_severity"] +
        0.16 * result["exposure"] +
        0.12 * (1 - result["accountability"]) +
        0.10 * (1 - result["detectability"]) +
        0.06 * (1 - result["inclusion_strength"])
    )

    result["governance_need"] = (
        0.30 * result["harm_severity"] +
        0.22 * result["exposure"] +
        0.18 * (1 - result["accountability"]) +
        0.16 * (1 - result["detectability"]) +
        0.14 * (1 - result["inclusion_strength"])
    )

    return result.sort_values("review_priority", ascending=False)

baseline = score_ethics(decisions)
print("Baseline ethical risk ranking:")
print(baseline)

np.random.seed(42)
n_simulations = 10000
records = []
highest_risk = []

for simulation_id in range(n_simulations):
    simulated = decisions.copy()

    for col in [
        "harm_severity",
        "probability",
        "exposure",
        "detectability",
        "accountability",
        "inclusion_strength"
    ]:
        simulated[col] = np.random.normal(
            loc=decisions[col],
            scale=0.08
        ).clip(0, 1)

    scored = score_ethics(simulated).reset_index(drop=True)
    highest_risk.append(scored.iloc[0]["design_decision"])

    for rank, row in scored.iterrows():
        records.append({
            "simulation_id": simulation_id,
            "design_decision": row["design_decision"],
            "ethical_risk": row["ethical_risk"],
            "review_priority": row["review_priority"],
            "governance_need": row["governance_need"],
            "rank": rank + 1
        })

simulation_df = pd.DataFrame(records)

risk_winners = (
    pd.Series(highest_risk)
    .value_counts(normalize=True)
    .rename("probability_highest_review_priority")
    .reset_index()
)

risk_winners.columns = ["design_decision", "probability_highest_review_priority"]
risk_winners["probability_highest_review_priority"] *= 100

rank_stability = (
    simulation_df
    .groupby("design_decision")
    .agg(
        mean_ethical_risk=("ethical_risk", "mean"),
        sd_ethical_risk=("ethical_risk", "std"),
        mean_review_priority=("review_priority", "mean"),
        mean_governance_need=("governance_need", "mean"),
        median_rank=("rank", "median"),
        mean_rank=("rank", "mean"),
        best_rank=("rank", "min"),
        worst_rank=("rank", "max")
    )
    .reset_index()
    .sort_values(["median_rank", "mean_rank"])
)

print("\nProbability each design decision has highest review priority:")
print(risk_winners)

print("\nEthical risk rank stability:")
print(rank_stability)

plt.figure(figsize=(10, 6))
plt.bar(risk_winners["design_decision"], risk_winners["probability_highest_review_priority"])
plt.xticks(rotation=20, ha="right")
plt.ylabel("Probability of highest review priority (%)")
plt.title("Ethical Design Review Priority Under Uncertainty")
plt.tight_layout()
plt.show()

baseline.to_csv("ethical_design_baseline_scores.csv", index=False)
risk_winners.to_csv("ethical_design_uncertainty_winners.csv", index=False)
rank_stability.to_csv("ethical_design_rank_stability.csv", index=False)
simulation_df.to_csv("ethical_design_simulation_records.csv", index=False)

This workflow is useful because ethical risk is often uncertain. Teams may underestimate harm, overestimate accountability, or miss exposure across affected groups. Simulation does not solve ethics, but it helps reveal when a design decision should receive stronger governance before moving forward.

Back to top ↑

GitHub Repository

The companion repository will provide a reproducible technical workspace for exploring the modeling, simulation, documentation, and implementation ideas associated with this article. The article folder is organized for multi-language ethics, power, inclusion, and design-thinking research and includes folders for Python, R, Julia, C++, Fortran, C, Rust, Go, SQL, notebooks, documentation, raw data, processed data, and outputs.

The repository structure is designed to support reproducible ethics and inclusion research rather than isolated code examples. The language-specific folders allow the same inclusion, burden, power, ethical risk, accessibility, accountability, and uncertainty logic to be explored across statistical, scientific, systems, and database workflows. The documentation and data folders help preserve assumptions, stakeholder definitions, risk criteria, governance safeguards, accessibility constraints, participation commitments, and repair pathways so that ethical design judgments remain traceable.

Back to top ↑

Conclusion

Ethics, power, and inclusion are not optional additions to design thinking. They are central to whether design thinking deserves to be called human-centered at all. A process that listens without accountability, includes without influence, prototypes without safeguards, or scales without repair may use the language of design thinking while reproducing the failures it claims to solve.

Ethical design thinking begins by naming power. It asks who defines the problem, who controls evidence, who makes decisions, who carries burden, who is excluded, and who can demand repair. It treats inclusion as more than representation. It treats accessibility as a foundation. It treats lived experience as knowledge. It treats participation as a question of influence, not just attendance.

It also asks design teams and institutions to be honest about limits. Not every problem should be solved by design methods. Some require rights, funding, governance, redistribution, legal protection, community control, or institutional reform. Ethical design thinking helps clarify where design can contribute and where broader change is required.

The strongest version of design thinking is not merely creative, empathetic, or iterative. It is accountable. It recognizes that every design redistributes possibility. Some people gain access, time, safety, dignity, and voice. Others may lose them. The work of ethical design is to make those distributions visible before they harden into systems.

Design thinking becomes serious when it asks not only “What can we make?” but “What should we make, who should decide, who may be harmed, how will we know, and how will we repair what goes wrong?”

Back to top ↑

Related articles

• What Is Design Thinking?
• Human-Centered Problem Solving
• Empathy and Stakeholder Research in Design Thinking
• Design Research Methods: Contextual Inquiry and Synthesis
• Problem Framing in Design Thinking
• Insight Generation in Design Thinking
• Prototyping in Design Thinking
• Testing and Validation in Design Thinking
• Iteration and Experimentation in Design Thinking
• Service Design in Design Thinking
• Design Thinking and Behavioral Design
• Design Thinking and Strategy
• Co-Design and Participatory Design
• Design Thinking in Public Policy

Further reading

• Costanza-Chock, S. (2020) Design Justice: Community-Led Practices to Build the Worlds We Need. Cambridge, MA: MIT Press. Available at: https://mitpress.mit.edu/9780262043458/design-justice/.
• Design Justice Network (2018) Design Justice Network Principles. Available at: https://designjustice.org/read-the-principles.
• Holmes, K. (2018) Mismatch: How Inclusion Shapes Design. Cambridge, MA: MIT Press. Available at: https://mitpress.mit.edu/9780262539487/mismatch/.
• IDEO (n.d.) Human-centered design. Available at: https://www.ideo.com/.
• Microsoft (n.d.) Inclusive Design Toolkit. Available at: https://inclusive.microsoft.design/.
• Norman, D.A. (2013) The Design of Everyday Things. Revised and expanded edition. New York: Basic Books. Available at: https://www.basicbooks.com/titles/don-norman/the-design-of-everyday-things/9780465050659/.
• Schuler, D. and Namioka, A. (eds.) (1993) Participatory Design: Principles and Practices. Boca Raton: CRC Press. Available at: https://www.routledge.com/Participatory-Design-Principles-and-Practices/Schuler-Namioka/p/book/9780805809510.
• United Nations (2006) Convention on the Rights of Persons with Disabilities. Available at: https://www.ohchr.org/en/instruments-mechanisms/instruments/convention-rights-persons-disabilities.

Back to top ↑

References

• Costanza-Chock, S. (2020) Design Justice: Community-Led Practices to Build the Worlds We Need. Cambridge, MA: MIT Press. Available at: https://mitpress.mit.edu/9780262043458/design-justice/.
• Design Justice Network (2018) Design Justice Network Principles. Available at: https://designjustice.org/read-the-principles.
• Holmes, K. (2018) Mismatch: How Inclusion Shapes Design. Cambridge, MA: MIT Press. Available at: https://mitpress.mit.edu/9780262539487/mismatch/.
• IDEO (n.d.) Human-centered design. Available at: https://www.ideo.com/.
• Microsoft (n.d.) Inclusive Design Toolkit. Available at: https://inclusive.microsoft.design/.
• Norman, D.A. (2013) The Design of Everyday Things. Revised and expanded edition. New York: Basic Books. Available at: https://www.basicbooks.com/titles/don-norman/the-design-of-everyday-things/9780465050659/.
• Schuler, D. and Namioka, A. (eds.) (1993) Participatory Design: Principles and Practices. Boca Raton: CRC Press. Available at: https://www.routledge.com/Participatory-Design-Principles-and-Practices/Schuler-Namioka/p/book/9780805809510.
• United Nations (2006) Convention on the Rights of Persons with Disabilities. Available at: https://www.ohchr.org/en/instruments-mechanisms/instruments/convention-rights-persons-disabilities.

Back to top ↑