Crisis Communications in Washington, DC: How Leaders Protect Trust Under Pressure

Executive brief

What should leadership decide before a crisis begins?

Six choices determine whether the response team can protect people, preserve an accurate record, and support the decisions that follow.

Washington context

Why is Washington, DC a distinct market for crisis communications?

A Washington crisis can move through government, national media, trade associations, advocates, investors, employees, and online communities within the same hour.

Washington concentrates institutions with the authority to investigate, regulate, legislate, litigate, convene, and amplify. A product incident can become a committee question. A personnel dispute can become a governance story. A leaked memo can affect coalition support. A cyber event can trigger operational, law-enforcement, regulatory, investor, employee, and customer communications at once. The response team therefore works across overlapping systems of accountability.

The city's professional ecosystem adds capacity and complexity. Law firms, public affairs consultancies, cybersecurity specialists, lobbyists, opposition researchers, pollsters, government relations teams, national newsrooms, and crisis public relations firms may all enter the same matter. Each discipline sees a legitimate portion of the problem. Leadership still needs a common record, a clear authority structure, and one set of priorities. Parallel workstreams become dangerous when they publish from different facts or interpret the same audience in incompatible ways.

Speed has changed as well. A reporter, employee, member, activist, regulator, investor, customer, or anonymous account can introduce new material into public view. Search engines index the response. Social platforms detach clips from context. News influencers translate the story for communities that may never encounter the original statement. Generative systems can summarize a disputed claim beside established reporting and official sources. A polished press statement reaches only one part of this environment.

This is why Washington organizations need a digital-first crisis system. Digital-first describes the operating reality, not a preference for social media. The system assumes that facts will be discovered, copied, questioned, searched, summarized, and recirculated across platforms. It gives the response team the ability to publish evidence, serve stakeholders, track interpretation, and preserve continuity while legal and operational decisions continue.

The role of a digital-first partner is consequently precise. The partner translates verified decisions into a coherent public record, connects audience intelligence to channel action, provides the production capacity to move quickly, and measures where understanding remains weak. Legal counsel owns legal interpretation. Incident leaders own operational decisions. Subject-matter experts own factual verification. Senior leadership owns risk acceptance. Communications makes those decisions legible to the people who need to act.

Triage

What qualifies as a communications crisis?

A difficult headline becomes a crisis when the issue threatens people, operations, legal standing, trust, mission, revenue, or the organization's ability to act.

Teams often waste the opening period debating the label. The productive question is whether the matter requires cross-functional authority and a faster decision cycle than normal business. An adverse article may remain a media-relations issue. A fast-moving article that prompts regulator questions, employee concern, customer cancellations, and false claims requires crisis command. The shift depends on consequence, velocity, uncertainty, and stakeholder reach.

Severity can rise or fall. A Level 1 rumor may disappear after direct correction. A Level 2 product complaint may reveal a larger safety problem. The plan should name escalation triggers: credible evidence of harm, government inquiry, litigation, executive involvement, employee walkout, customer disruption, coordinated manipulation, rapid media acceleration, a material disclosure question, or a marked change in audience behavior.

The trigger language should be observable. Terms such as “significant attention” invite debate. A stronger rule names the condition: three national reporter inquiries in one hour, a formal regulator request, a top-ten customer escalation, a 30 percent increase in service calls, or an unverified video crossing a defined reach threshold. Thresholds will vary, but the discipline makes activation faster. FEMA's National Incident Management System provides broader support for shared terminology, scalable structures, and coordinated incident management.^[6]

Original framework

What is the Digital Crisis Response System?

The Digital Crisis Response System is a six-layer operating model that turns verified decisions into audience understanding and useful action.

Most crisis plans inventory scenarios, contacts, and draft statements. Those materials are valuable. They rarely explain how intelligence becomes a decision, how a decision becomes a digital asset, how that asset reaches affected people, or how the team knows whether the response worked. The six layers below create that connection.

The model follows a single operating rule: every communications action should originate in an authorized decision, serve a named audience need, use the most direct credible channel, and produce a signal the team can interpret. This rule prevents activity from becoming a substitute for judgment.

The CDC's Crisis and Emergency Risk Communication framework offers a complementary standard for emergency communicators. Its six principles are to be first, be right, be credible, express empathy, promote action, and show respect.^[1] These principles remain useful beyond public health. A fast response establishes presence. Accuracy protects the record. Credibility depends on consistency and proof. Empathy recognizes the people carrying the consequence. Action gives the audience a productive next step. Respect shapes tone, access, and follow-through.

Rapid response

What should happen during the first hour, first four hours, and first day?

The opening period should reduce uncertainty, protect affected people, establish a reliable source, and create a sustainable decision rhythm.

A holding statement should acknowledge the situation, describe the organization's current action, state where verified updates will appear, and avoid conclusions that the evidence does not support. It should sound like the organization speaking under pressure, not a template. If people face an immediate safety, service, privacy, or financial consequence, the action guidance belongs near the beginning.

The response team also needs a strict rhythm. A typical Level 3 crisis may use a 15-minute intelligence review, a 30-minute decision meeting, and a 60-minute publication cycle during the opening period. The exact intervals follow the incident. What matters is the separation of functions. Intelligence gathers and verifies. Command decides. Communications publishes. The recorder preserves the decision and its basis.

Scheduled social posts, marketing email, paid campaigns, event promotions, executive content, and fundraising messages deserve immediate review. A blanket pause can be appropriate, although it may also interrupt essential service communications. The channel owner should assess relevance and potential harm, then record the decision.

Reputation and brand management

How should an organization manage reputation before, during, and after a crisis?

Reputation is the accumulated judgment that stakeholders make about conduct, competence, reliability, and relevance. Communications can clarify that judgment. Conduct supplies its foundation.

Brand standards still matter under pressure. The voice, visual system, executive behavior, and customer experience should remain recognizable. Rigidity creates risk, however. A campaign slogan may sound evasive when people need an operational answer. Highly produced video can feel distant when a direct executive explanation is warranted. The crisis team should preserve the organization's character while adjusting pace, detail, and tone to the human stakes.

Preparation begins with a reputation baseline. Which audiences grant the organization permission to act? Which promises carry the greatest expectation? Where do perception and performance diverge? What prior issues remain searchable? Which executives, experts, customers, members, community leaders, employees, and partners can speak from direct knowledge? Ipsos frames reputation research around the issues and actions that build trust, including gaps between what organizations say and what they do.^[14] That gap analysis gives the crisis plan substance.

During the crisis, the organization should avoid treating sentiment as reputation. Online volume may reflect a small, highly active group, automated behavior, media attention, genuine stakeholder harm, or a mix of all four. Leadership needs to know who holds the view, why it matters, whether behavior is changing, and what evidence could change the conclusion. A negative post count cannot answer those questions.

After the peak, recovery moves through three records. The operational record describes what changed. The stakeholder record shows who received support and what remains unresolved. The public record includes media coverage, social posts, search results, official documents, executive statements, AI-generated summaries, and third-party analysis. All three need attention. A corrected process with a stale public record leaves future audiences with an incomplete account.

Edelman's 2026 Trust Barometer surveyed 33,938 respondents across 28 countries and described audiences retreating toward smaller, more familiar circles of trust.^[11] That global finding should not be treated as a forecast for every Washington stakeholder. It supports a practical implication: institutional statements need reinforcement from credible people and networks close to the audience. The appropriate messenger might be a local leader, technical expert, employee, member, physician, customer, regulator, or community partner. The choice requires evidence and consent.

Opinion research

How can opinion research improve crisis decisions?

Research is valuable when it identifies who believes what, what evidence shapes that belief, and which action would reduce harm or restore confidence.

A crisis dashboard contains abundant behavioral data. Search queries show information demand. Web paths show where people seek help. Call-center logs reveal friction. Social and media analysis show themes and messengers. These sources are fast, yet none represent public opinion by themselves. A well-designed research program adds questions that passive monitoring cannot answer.

The first question is decision fit. A rapid qualitative interview may be the right tool for understanding why employees distrust a proposed message. A probability-based survey may be necessary for estimating public attitudes. A stakeholder panel can track a defined professional community. An opt-in sample can support directional message testing when the limitation is explicit. Pew Research Center's polling guidance explains why representative inference depends on sampling, while its methods research warns that online nonprobability samples vary in recruitment, selection, weighting, and data quality.^[15]

Question wording deserves executive attention. Terms such as “crisis,” “failure,” “cover-up,” “attack,” and “misinformation” can introduce assumptions. Ask a factual exposure question before an evaluative question. Rotate choices where appropriate. Preserve a neutral control when testing response language. Publish the field dates, sample, recruitment method, weighting, question wording, and margin of sampling error where one applies.

Research timing also matters. A baseline conducted before a crisis reveals existing trust and vulnerability. An immediate pulse maps exposure and urgent information needs. A later wave measures whether corrective action reached the intended audience. Long-term tracking shows whether reputation recovered or simply disappeared from daily conversation. The team should resist commissioning a poll to validate a preferred message. The purpose is to identify reality early enough to change the response.

Media response

How should a crisis team work with journalists?

A reporter needs a reachable source, a verified record, a clear explanation, access to the right expert, and an honest account of what remains unresolved.

Washington media relations often begins before a crisis. Beat knowledge, subject-matter access, responsive press operations, and accurate background material establish whether the organization can function as a reliable source. A relationship cannot secure favorable coverage. It can make the exchange faster and reduce avoidable confusion when the facts are moving.

The media team should classify each inquiry. Is the journalist confirming a fact, seeking reaction, building a broader accountability story, requesting technical context, or working from material the organization has not seen? What is the deadline? Which outlet and audience are involved? Has the reporter already published? Which claim carries the greatest factual consequence? This classification informs the response path and the right spokesperson.

A useful press packet may include the latest statement, a time-stamped fact sheet, executive and expert biographies, a plain-language chronology, downloadable images or video, data definitions, contact details, and links to primary documents. Cision's 2025 survey of more than 3,000 journalists found that 72% named press releases as the most useful resource public relations teams can provide, while 86% said they would reject a pitch that did not align with their beat or audience.^[13] The finding favors relevance and utility over indiscriminate outreach.

The 2026 edition, based on more than 1,800 journalists worldwide, reported that 50% identified misinformation as a leading challenge.^[12] Vendor research and global samples require appropriate caution. The operational lesson is reasonable: reporters under verification pressure benefit from source documents, named experts, precise corrections, visual evidence, and clear provenance.

Corrections should target material facts. Contact the reporter with the exact statement at issue, the evidence supporting the correction, and an accessible source. Avoid turning a defensible disagreement over framing into an allegation of factual error. If the organization made the mistake, correct its own record first and notify affected reporters promptly.

Spokesperson preparation should cover substance, authority, and pressure. The spokesperson needs to understand the incident, the public consequences, and the limits of current knowledge. They should know which commitments the organization can make. Practice hostile and sympathetic questions, interrupted answers, false premises, silence, live remote interviews, hallway encounters, and the handoff to technical experts. Media training conducted only in a quiet conference room leaves important conditions untested.

The organization should also decide when an executive belongs in front. Chief executives convey accountability and direction. Technical leaders may carry greater credibility on cause and remedy. Local leaders may understand community impact. Affected employees or customers should never be used as reputation props. Their participation requires informed consent, support, and a genuine reason for inclusion.

The public record

How should websites, search, social platforms, and AI answers support the response?

The organization needs a canonical, accessible, current source that people and machines can retrieve while the story changes around it.

The crisis page should use a durable URL, a clear title, a visible date and update time, direct language, descriptive subheads, contact and assistance pathways, and links to supporting documents. Major changes belong in a dated log. Superseded information should be marked rather than silently removed when the historical record matters. The page needs to work on mobile devices, load quickly, meet accessibility requirements, and remain available under traffic pressure.

Search demand can reveal what the public cannot find. Queries combining the organization with words such as breach, recall, safety, lawsuit, resignation, closure, refund, scam, or investigation show the questions forming around the issue. The team should answer verified, material questions on the canonical page and within targeted stakeholder content. Search advertising may help affected people find official assistance when harmful or irrelevant results dominate. Any paid response requires precise targeting, legal review, clear sponsorship, and continuous monitoring.

Social channels have three jobs: point people to the authoritative record, deliver urgent action in the formats audiences use, and receive questions or reports that the command team must see. The channel manager should avoid debating every critic. Responses should prioritize safety, factual consequence, service, and reach. A visible claim with limited consequence may deserve monitoring. A low-volume claim that directs customers to a fraudulent site demands immediate action.

The American news environment is fragmented. Pew Research Center found in late 2025 that 70% of U.S. adults trusted information from local news organizations at least some, 56% said the same of national news organizations, and 37% trusted information from social media sites at least some.^[7] These measures do not assign credibility to a specific outlet or post. They show why one-channel distribution leaves predictable gaps.

Generative search adds another surface. Google says its AI features use information from the Search index and that the same foundational SEO practices remain relevant. Pages must be indexable and eligible to display a snippet. Google also advises website owners to keep important information in text, use clear internal links, support a good page experience, and ensure structured data matches visible content.^[16]

For crisis teams, this means basic publishing discipline matters more than speculative optimization. Place the confirmed facts in crawlable text. Use names, dates, locations, products, programs, and source documents consistently. Link from the newsroom and relevant service pages. Update old pages that could confuse the record. Monitor search results and major AI answers for consequential inaccuracies, then strengthen the public evidence available to correct them.

Generated answers may remain wrong after the source page changes. The response team should capture the query, date, answer, cited sources, and harm. It can then assess whether the problem comes from the organization's page, a third-party source, an indexing delay, entity confusion, or model behavior. The remedy may involve correcting owned content, requesting a publisher correction, strengthening third-party evidence, updating business or organization profiles, or using the platform's feedback mechanism. A promise of immediate AI-answer control would be misleading.

False narratives and impersonation

How should organizations respond to rumors, manipulated media, and impersonation?

The response should follow consequence and audience exposure. Broad denial can amplify an obscure claim, while silence can leave a harmful fraud or false instruction unchallenged.

The first task is classification. A rumor may come from misunderstanding. Misinformation describes false information shared without necessarily establishing intent. Disinformation involves deliberate deception. Malinformation uses genuine material in a harmful or misleading context. Impersonation can support fraud, harassment, market manipulation, reputational attack, or access to sensitive systems. The terminology matters because intent, evidence, legal options, and response channels differ.

The FBI has warned that malicious actors use text and AI-generated voice messages to impersonate senior U.S. officials. Its May 2025 notice described campaigns seeking access to personal accounts and contacts, while a later update said similar activity had occurred since at least 2023.^[17] The FTC's review of AI-enabled voice cloning identifies intervention points before, during, and after misuse, including authentication, real-time monitoring, detection, and post-use evaluation. It also concludes that no single intervention solves the problem.^[18]

An organization should establish verification norms before an attack. Publish the official domains, accounts, executive profiles, media contact, payment practices, and process for confirming unusual requests. Use multifactor authentication and strong internal escalation procedures. Maintain high-quality reference audio and video where appropriate. Record executive travel and event appearances with care. Train staff to verify urgent requests through a known second channel.

A rumor-control page can help when the same false claims recur. FEMA uses dedicated rumor-response pages during disasters and warns that rumors and misleading information can spread quickly.^[19] The format works when the page answers a real public need, carries clear update dates, cites primary information, and gives people a safe action. It becomes counterproductive when it repeats every provocation or introduces claims that few people had encountered.

Language should protect the affected audience from repetition effects. Lead with the verified fact. Name the false claim only as much as necessary to identify it. Explain how the organization knows. Provide the correct action. Repeat the official verification pathway. If the team cannot confirm whether an artifact is synthetic, describe it as unverified and explain the verification process.

Cyber and operational incidents

How should communications integrate with cyber incident response?

Cyber communications should operate inside enterprise risk management, with security, legal, privacy, operations, leadership, and communications working from the same incident record.

NIST Cybersecurity Framework 2.0 gives organizations a common language for managing cybersecurity risk and explicitly includes communication inside and outside the organization. The framework connects cybersecurity with financial, privacy, reputational, and other enterprise risks.^[2] NIST Special Publication 800-61 Revision 3, issued in April 2025, further integrates incident response across cybersecurity risk management to improve detection, response, and recovery.^[3]

This integration changes the communications brief. The team needs accurate descriptions of affected systems, service availability, data exposure, containment, restoration, investigation, customer action, employee guidance, third-party involvement, and regulatory status. Each term carries technical and legal meaning. Communications should maintain a plain-language translation layer while preserving the expert definition beneath it.

Public companies face specific disclosure duties. The SEC requires disclosure of a material cybersecurity incident under Item 1.05 of Form 8-K within four business days after the company determines the incident is material. The deadline runs from the materiality determination, and that determination must occur without unreasonable delay.^[4] This rule does not prescribe every customer, employee, media, law-enforcement, or state notification. Counsel should identify the obligations for the organization and incident.

The communications team should avoid premature certainty. “No evidence of access” differs from “no access occurred.” “Systems are operational” differs from “all services are fully restored.” “We contained the activity” does not establish the root cause. A fact library that pairs approved public language with the underlying technical definition reduces accidental overstatement.

CISA provides incident and vulnerability response playbooks and tabletop exercise packages that organizations can use to rehearse information sharing, response, and recovery.^[5] The communications module should be concrete. Can the team publish if the main content management system is unavailable? Are media lists accessible outside the network? Can leadership join a secure call? Who monitors a fraudulent support account? Can the organization send an accessible alert if customer records are offline?

Performance intelligence

How should leaders measure crisis communications?

The dashboard should show whether people received accurate information, understood it, trusted the response enough to act, and experienced the promised remedy.

Media volume and social sentiment offer context. They do not establish effectiveness. A response can produce negative coverage while protecting customers through clear action. It can also generate favorable commentary while leaving employees uninformed. Measures need to follow the stakeholder and the decision.

The team should establish a baseline for the measures that matter before a crisis. Normal media volume, branded search, site traffic, customer contacts, employee questions, trust, and issue awareness provide the comparison. Without a baseline, a spike may look dramatic while revealing little about consequence.

Sentiment analysis requires human review. Sarcasm, technical language, political speech, coordinated behavior, and culturally specific usage can mislead automated classification. Report the method and confidence. Separate the views of affected customers from general commentary. Distinguish original posts from reshares. Use qualitative examples to explain the numbers without selecting only the most favorable evidence.

Every metric should map to an owner and a decision. Time to first response can change approval rules. Repeated customer confusion can change the FAQ and call-center script. Low employee comprehension can change manager briefings. A stale search result can trigger a publisher correction or owned-content update. A dashboard that never changes action becomes decoration.

Scenario planning

What should four common Washington crisis plans include?

The operating model stays consistent. The evidence owners, stakeholders, obligations, channels, and proof of recovery change with the scenario.

A tabletop exercise should force decisions. Give the team incomplete facts, a competing executive priority, a reporter deadline, an inaccessible system, an employee leak, and an emerging false claim. Observe who calls the meeting, who verifies the record, how legal and communications resolve uncertainty, and whether the digital team can publish. The gaps become a funded readiness plan.

Gigawatt Group's role

Where can a digital-first partner strengthen crisis communications?

Gigawatt Group can connect public affairs strategy, stakeholder intelligence, creative production, digital publishing, targeted distribution, and performance reporting inside the client's crisis command structure.

Gigawatt Group's existing public affairs and strategic communications practice combines campaign planning, audience and stakeholder mapping, messaging, integrated communications, data insight, targeted media, and reporting. Those capabilities become especially valuable when a crisis compresses the time between evidence, decision, and public action.

The firm's role should match the incident and the client's existing team. Some organizations need a readiness program: vulnerability mapping, channel audit, audience research, message architecture, scenario planning, content templates, dashboards, and exercises. Others need an embedded digital workstream during a live matter. A client may already have lead crisis counsel, a national public relations firm, government affairs advisers, cybersecurity incident responders, or an internal command center. Gigawatt can provide the connective digital layer rather than duplicating those roles.

Gigawatt's broader operating model supports the work from business and market intelligence through strategy, content infrastructure, integrated execution, performance intelligence, and optimization. The structure matters in a crisis because strategy without production arrives late, while production without intelligence creates noise. A connected partner can move from stakeholder question to approved asset to measured response within the same system.

Digital creative is another source of leverage. Complex incidents often require a service-status interface, explainer graphic, executive video, accessible FAQ, customer email, policymaker brief, paid search message, social card, press asset, and internal manager kit. These assets should share a factual core while serving different use cases. Gigawatt's integrated creative and media resources can reduce handoffs when time and consistency are both critical.

The engagement should begin with boundaries. Legal advice remains with counsel. Law enforcement, cybersecurity forensics, emergency management, regulatory judgment, and executive authority remain with their designated owners. Gigawatt can work beside those functions to make verified decisions understandable, findable, actionable, and measurable. That is the digital-first contribution to crisis command.

Implementation

What should leaders build during the next 90 days?

A focused readiness program can establish authority, evidence, publishing capacity, audience intelligence, and rehearsal without pretending every scenario can be predicted.

The final test is practical: can a qualified alternate open the evidence record, convene the right people, publish through an approved channel, and brief leadership if the primary owner is unavailable? A plan that depends on one person is a vulnerability.

Frequently asked questions

Questions Washington leaders ask about crisis communications

Research record

Sources and interpretation notes

The sources below support the factual claims in this report. Gigawatt Group's frameworks, recommendations, and service positioning are original interpretations.

1. Centers for Disease Control and Prevention. Crisis and Emergency Risk Communication Manual, accessed August 13, 2026. Used for the six emergency communication principles. The framework was developed for public health emergencies; this report applies the principles more broadly and labels that adaptation.
2. National Institute of Standards and Technology. The NIST Cybersecurity Framework 2.0, February 2024. Used for enterprise cybersecurity risk and internal and external communication.
3. National Institute of Standards and Technology. SP 800-61 Revision 3: Incident Response Recommendations and Considerations for Cybersecurity Risk Management, April 2025. Used for integrated incident response and recovery.
4. U.S. Securities and Exchange Commission. Cybersecurity Risk Management, Strategy, Governance, and Incident Disclosure, accessed August 13, 2026. Used for Item 1.05 Form 8-K timing. Organizations should obtain legal advice for their circumstances.
5. Cybersecurity and Infrastructure Security Agency. CISA Tabletop Exercise Packages and Planning, Response, and Recovery, accessed August 13, 2026. Used for exercise and response-readiness guidance.
6. Federal Emergency Management Agency. National Incident Management System, accessed August 13, 2026. Background for scalable incident coordination and common operating structures.
7. Pew Research Center. How Americans' trust in information from news organizations and social media sites has changed over time, October 29, 2025. Survey of 3,445 U.S. adults conducted October 13–19, 2025. Used for the source-trust chart.
8. Pew Research Center. Social Media and News Fact Sheet, September 25, 2025. Survey conducted August 18–24, 2025. Used for U.S. social news consumption.
9. Pew Research Center. America's News Influencers, November 18, 2024. Used for U.S. news-influencer consumption, including the age comparison.
10. Gallup. Trust in Media at New Low, October 2, 2025. Used for the U.S. mass-media trust measure. Gallup's wording and trend definition differ from Pew's source-specific measures.
11. Edelman. 2026 Edelman Trust Barometer, 2026. Global online survey of 33,938 respondents in 28 countries, fielded October 25–November 16, 2025. Used for the finding on insularity and familiar trust circles. It is not a Washington-specific estimate.
12. Cision. 2026 State of the Media Report and five topline findings, 2026. Global survey of more than 1,800 journalists. Used for the misinformation challenge measure. Cision is a communications software provider.
13. Cision. 2025 State of the Media Report findings, May 7, 2025. Global survey of more than 3,000 journalists. Used for the relevance and press-release findings.
14. Ipsos. Corporate Reputation, accessed August 13, 2026. Used for reputation, trust, expectation, and the gap between statements and conduct. Ipsos provides reputation advisory services.
15. Pew Research Center. Public Opinion Polling Basics and Comparing Two Types of Online Survey Samples, accessed August 13, 2026. Used for sampling and method cautions.
16. Google Search Central. AI Features and Your Website and Optimizing Your Website for Generative AI Features on Google Search, accessed August 13, 2026. Used for search, AI feature, technical, and visible-content guidance.
17. Federal Bureau of Investigation. Senior U.S. Officials Impersonated in Malicious Messaging Campaign, May 15, 2025, and updated notice, December 19, 2025. Used for AI-generated voice impersonation risk.
18. Federal Trade Commission. Approaches to Address AI-enabled Voice Cloning, April 8, 2024. Used for intervention points and the limits of single-solution approaches.
19. Federal Emergency Management Agency. Hurricane Rumor Response, accessed August 13, 2026. Used as an example of a dedicated official rumor-response hub.

Editorial note: Statistics from different organizations use different samples, dates, geographies, questions, and methods. They appear as separate evidence points and should not be combined into a single trend calculation. Legal, regulatory, cybersecurity, emergency-management, polling, and platform decisions require qualified advisers and case-specific judgment.