Making Resilience Count: Translating Risk Reduction into Investment Signals

How do we turn insights about how resilience reduces risk into evidence that investors can act on? Explore findings from an April 22, 2026, workshop organized by Duke University during DC Climate Week.

Communities know where flooding is likely to occur. Insurers know which buildings are likely to fail. Engineers know what it would take to keep the lights on. The US Chamber of Commerce estimates a 6-to-1 return on resilience investment—some analyses put it at 13-to-1. And still, year after year, the investment does not happen at the scale the risk demands.

The knowledge exists. The project plans exist. The money exists. What does not yet exist is a shared language that allows these pieces to connect with each other in a way that enables more consistent decision-makers across sectors, companies, and communities.

This is not a new problem. For more than a decade, researchers, practitioners, and resilience initiatives have been working to connect risk science to investment decisions—to turn what engineers, insurers, community planners, and financial analysts each know into evidence that the others can act on. The work has produced material insights. It has not yet produced the shared translation architecture.

The same translation failures identified in 100 Resilient Cities early measurement work and in RAND and Rockefeller Foundation's efforts to value the "resilience dividend" still show up in every room where this conversation happens: a project that clearly reduces risk, described in terms that no lender can underwrite; a community that needs protection, unable to articulate its value in the language of capital markets; an insurer with 30 years of claims data that hasn't yet changed a single lending rate.

The cost of this gap is not abstract. It shows up in which communities get funded and which don't, in which infrastructure gets built before the disaster and which gets rebuilt after, and in who can afford to wait—and who can't.

Duke University's RESILE convened "Making Resilience Count: Translating Risk Reduction into Investment Signals" as part of DC Climate Week because this translation problem sits at the center of its mission. The session brought together practitioners from insurance, capital markets, development finance, municipal government, engineering, community resilience, and research, not to produce another framework, but to pressure-test where the connections are forming and where they keep breaking. This summary captures the workshop's findings, opportunities, and action agenda. The central takeaway is that resilience investment will not scale through better metrics alone; it requires practical handoffs that link risk reduction to performance, capital, insurance, public budgets, and community outcomes.

This workshop and these proceedings use the definition established in FEMA's National Resilience Guidance (2024): resilience is the ability to prepare for threats and hazards, adapt to changing conditions, and withstand and recover rapidly from adverse conditions and disruptions. Even with that shared definition in hand, the translation problem persists because the challenge is not what resilience means, but what it is worth, to whom, and how that value moves across sectors.

Why Resilience Value Still Does Not Translate into Investment

The workshop surfaced a consistent pattern: the problem is not a lack of data, expertise, or activity. Across sectors, useful measures already exist, but they were built for different decisions and do not yet connect in ways that change financing, insurance, public investment, or community outcomes. Four challenges stood out: sector-specific measures are not yet interoperable; avoided loss is necessary but insufficient; outputs are often mistaken for outcomes; and better metrics cannot fix weak project structure.

Useful Measures Are Being Built in Separate Lanes

The workshop clarified that the resilience valuation challenge is not a lack of activity, data, or expertise. Across sectors, significant progress is already being made. Insurers, lenders, engineers, governments, investors, and community practitioners are all developing useful ways to measure resilience, risk reduction, and value. The challenge is that these measurements are usually built for different decisions, time horizons, definitions of value, and accountability systems. Each measure may be useful within its own field, but it is not automatically interoperable with others.

As a result, the same project can look different depending on who is making the decision. An engineer may see risk reduction, measured through flood-depth reduction, asset condition, restoration time, criticality, or system performance. A community may see protection, reflected in displacement risk, mental health, adaptive capacity, or access to basic services. For an insurer, the relevant question may be whether the project lowers claims exposure, claim severity, coverage constraints, or loss ratios. A local government may focus on reduced exposure, improved public health, lower capital needs, or more reliable service delivery. A green bank or lender may evaluate the same project through credit quality, collateral, cash flow, tenant stability, and downside risk.

The same translation challenge holds true even for basic terms like asset and avoided loss. For a community, an asset may be a neighborhood center, a trusted elder, or a social network. For an engineer, investor, insurer, or banker, an asset has a technical or financial meaning. Avoided loss means avoided disaster costs in a Federal Emergency Management Administration (FEMA) context, revenue protection and operational continuity in a corporate context, and credit risk and default probability in a lending context.

This is not a vocabulary problem—it genuinely reflects different decision contexts, time horizons, and accountability structures. Public budgets run on annual or political cycles. Capital markets operate on shorter return horizons. Infrastructure assets last decades. Household vulnerability can be immediate. Resilience benefits are long-term, probabilistic, and event-driven.

Participants did not describe a blank slate. They pointed to places where the translation from resilience value to decision-ready evidence is already beginning to happen, but usually within a specific sector, jurisdiction, or use case:

• Washington, DC's, stormwater credit trading program turned a regulatory requirement into a price floor and a tradable credit, making stormwater retention measurable and financeable within a single jurisdiction.
• The Insurance Institute for Business & Home Safety (IBHS) has nearly 100,000 FORTIFIED designations and decades of claims data linking structural standards to storm outcomes.
• RiskThinking.ai maps probability distributions of forward-looking physical risk at the asset level, informing probability of default and capital reserve questions at major financial institutions.
• The Nature Conservancy has developed a methodology to quantify flood-depth reduction and wave attenuation from coastal wetlands, expected to be accredited through Verra's Sustainable Development Verified Impact Standard (SD VISta) program.
• Montgomery County Green Bank underwrites resilience as a credit question, including downside protection, cash-flow stability, tenant retention, stress testing.

Together, these examples show that the building blocks are emerging for measuring and communicating the value of resilience more precisely. The challenge is that they remain fragmented: each translates resilience value for a particular decision, but the field still lacks a reliable way to connect those signals across sectors. Participants emphasized that this challenge would not be solved by agreeing on definitions or by creating one universal metric, whether for risk reduction specifically or for resilience more broadly.

Avoided Loss Is Necessary, but Not Enough to Move Capital

For more than a decade, the standard argument for resilience investment has been avoided loss: if a project reduces future damages, claims, emergency costs, or recovery costs, that value should be counted. The argument is correct. The 6-to-1 return on resilience investment is genuine. But avoided loss still does not move capital at the scale needed, because it has structural limits that better data alone cannot fix.

Avoided loss is probabilistic and depends on assumptions about future hazards that different audiences will dispute. It requires a counterfactual—what would have happened without the investment—that is always contestable. It often benefits multiple parties rather than the entity paying for the project, which makes it hard to build a repayment structure around. And it values something that did not happen, which makes it difficult to use in budgeting, credit, insurance, or political decisions where decision-makers need to show what they got, not what they prevented.

The avoided-loss frame has produced good research and important policy wins. But after a decade, it is clear that avoided loss cannot carry the full weight of the resilience investment case on its own. The more useful question is not only what damage was prevented, but what kept working. Avoided loss is necessary, but it is not sufficient as a shared investment signal.

The limits of avoided loss point to a broader measurement problem: many resilience metrics still show what was done, not what changed.

Outputs are not Outcomes

Once avoided loss is no longer enough to carry the investment case on its own, a third challenge becomes easier to name: some metrics can make progress look more complete than it is. The issue is not one bad metric. It is a recurring pattern of treating activity, coverage, or spending as evidence that resilience has improved. For example, counting the percentage of assets that have been hardened is not very useful unless those assets are also connected to the hazards they face, how essential they are to the system, and who would be most affected if they failed. Many dashboards have the same problem: they track activities completed rather than whether risk was reduced or people and services were better protected.

A persistent failure across sectors is treating what gets done as evidence of what changed. Percent of playgrounds with shade is not heat illness rates reduced. Miles of green infrastructure installed is not stormwater retained. Assets hardened is not outages prevented. Dollars spent is not displacement avoided. An organization can fill an implementation dashboard and still have made no measurable difference in resilience outcomes.

Table 1. Differences Between Outputs and Outcomes

This matters for investment because risk reduction has to become a financial signal before it can reliably affect capital decisions. Lenders, insurers, investors, budget offices, and rating agencies need more than proof that an activity occurred. They need evidence that the activity changed risk, reduced losses, improved service continuity, stabilized cash flow, protected asset value, or lowered future exposure. Without that connection, resilience progress may be real, but it remains difficult to price, underwrite, finance, or reward.

The gap between outputs and outcomes is not just analytical; it is also practical. Implementation dashboards are often designed to meet reporting requirements, communicate progress, and maintain accountability to the public. That makes output indicators such as trees planted, federal dollars captured, projects completed, or assets covered easier to track and easier to report than outcomes. But outcome indicators are what show whether losses declined, services continued, health risks fell, or financial risk changed.

This same logic should guide how dashboards and indicators are designed. A dashboard for public accountability is different from one used to assess credit risk, underwrite insurance, or inform insurance pricing. If a measurement tool tries to serve every audience at once, it may not serve any of them well.

The design question is: Which decision should this indicator inform, for whom, and what outcome would show that resilience actually improved?

But even outcome metrics only solve part of the problem. A stronger measure can show that resilience value exists; it cannot, by itself, determine whether a project has the authority, repayment structure, borrower, or capital fit needed to move forward.

Better Metrics Will Not Make a Project Financeable

This is where measuring value and preparing a project for financing have to be separated. A project can be valuable without being ready for financing. It can qualify for grants without fitting the requirements of private capital. It can reduce risk without generating a repayment stream. The practical distinction is between a project that produces good resilience outcomes and a borrower that can support a loan. Treating those as the same can create confusion about what problem needs to be solved.

This is why better metrics, on their own, cannot make a project ready for financing. Resilience investments often create benefits for many parties at once, while the cost sits with one owner or public agency. Reduced physical risk may also matter for credit, but lenders do not yet consistently translate that risk reduction into better financing terms. And if resilience needs are identified too late in the project cycle, they cannot shape scope, budget, repayment, or risk allocation.

These are not failures of value. They are failures of translation, timing, structure, authority, or capital fit. Naming the specific failure is often the first step toward fixing it.

Table 2. Resilience Term Definitions

The Opportunities

The challenges above do not point to one missing metric. They point to places where existing evidence can be made more useful for decisions. The opportunities that follow focus on signals that are already emerging across sectors—performance under stress, criticality, credit risk, and insurance outcomes—and could help translate measured resilience value into financing, insurance, public investment, and community decisions.

Use Performance Under Stress as a Shared Signal

From the workshop discussions, the resilience measure that is the clearest candidate for this type of cross-sector effort is performance under stress: whether systems, services, and community functions continue to operate during a disruption. This includes downtime avoided, outage duration reduced, time to recovery, service uptime, displacement avoided, revenue maintained during disruption, and claim frequency reduced.

What makes performance under stress powerful is that it speaks to every sector. Downtime is an engineering metric, but it is also a business interruption issue, a household stability question, a public finance signal, and a credit concern all at once. Participants noted that a single lost workday can be catastrophic for one family and irrelevant to another. Metrics need to be contextualized to the population they affect, not just the asset.

Displacement made the cascading logic especially clear. It is not only a human outcome; it connects to business interruption, tax revenue decline, healthcare costs, education systems, housing inflation in receiving communities, and property values. System failures in power, transportation, water, childcare, and communications show up simultaneously as lost workdays for households, production shutdowns for companies, and credit pressure for municipalities. These are the same failures, visible to every sector, but rarely measured as a shared signal.

If performance under stress is the shared signal, then the next question is what has to be protected for systems to keep functioning. That is where herd immunity and criticality become important.

Measure System Protection, Not Just Assets Protected

Two insights from the workshop point to better ways of measuring performance under stress.

The first is herd immunity: if perimeter assets in a community are protected, interior assets may benefit too. This is the same logic that applies to wildfire zones, where outer-ring hardening can protect inboard properties, and to flood systems, where perimeter protection determines what stays on the dry side. Protection is not additive, parcel by parcel, but systemic. A house built above code can still be damaged or destroyed if surrounding structures, infrastructure, or protective systems fail.

The second is criticality: not all protected assets matter equally to system performance. It is less cost-effective to harden assets in wealthy or low-risk areas while leaving critical assets exposed. Protecting a facility, road, pump station, school, childcare center, clinic, or substation that supports essential services or vulnerable populations may matter more than protecting a larger number of lower-risk assets.

Together, herd immunity and criticality show why traditional percentage metrics can be misleading. Measures such as percent hardened, percentage of assets protected, and other coverage-style indicators are used in transportation resilience planning, utility wildfire mitigation, climate adaptation reporting, infrastructure asset management, and public-sector portfolio reporting. These types of metrics often assume that resilience can be calculated by comparing protected assets to total assets. For example, a portfolio can show strong protection rates because assets in affluent or more secure areas were hardened, while higher-risk or operationally critical assets remain exposed. The metric rewards what was easy to protect, not what most needed protecting, and can obscure the more important question: which assets, people, services, and system functions are protected under stress?

Resilience is not simple addition. Some protections create benefits beyond the asset itself, and some assets matter far more because of the people, services, and systems that depend on them. A more accurate measure asks not only what was hardened or protected, but what functions were preserved, who benefited, and what continued operating during disruption.

Once resilience is measured in terms of protected functions and avoided disruption, the next question is whether those benefits are recognized in financial decisions. Credit risk is one place where that connection is beginning to emerge, but is still underdeveloped.

Use Credit Risk to Show the Cost of Delay

Credit risk offers an important opportunity for local governments and their finance partners to make the financial value of resilience more concrete. Workshop participants emphasized that physical risk reduction can matter for credit because unmanaged risk can weaken local tax bases, raise borrowing costs, and increase fiscal stress over time. The evidence points in the same direction across multiple hazards: heat exposure has been linked to higher municipal borrowing costs; sea-level-rise risk is already showing up in bond pricing where repayment depends on local real estate taxes; and wetland loss can raise bond yields by increasing flood and water-quality risk, especially in tax-reliant communities. Broader climate-finance analysis reaches the same conclusion: physical climate risks can raise public costs, reduce revenues, erode property tax bases, and increase debt and borrowing costs.

Corpus Christi, TX, offers a recent example of this dynamic. The issue was not a bond default, but a series of credit-rating actions tied to water-supply risk. Rating agencies downgraded or revised the outlook on city and utility debt as drought conditions, reservoir depletion, and the need for costly new water-supply projects raised concerns about future costs, revenues, and debt service coverage. The case shows how unmanaged physical risk can become a credit issue before a community ever reaches default.

Evidence from the mortgage market also shows that homes built to modern building codes are less likely to become 90+ days delinquent after a hurricane, and that defaults in modern-code homes are more likely to resolve favorably. This supports the broader point that physical risk reduction can affect credit risk at the asset level, while municipal bond research shows a parallel relationship between physical risk exposure and public-sector borrowing costs.

Workshop participants also highlighted emerging housing-market research that shows natural infrastructure can protect asset value: homes near mangroves in coastal Florida experienced smaller post-hurricane price declines, with proximity to mangroves preserving an estimated $20,000 to $40,000 per $1 million in home value, and up to $60,000 in some cases.

The opportunity is to frame that connection clearly. Resilience investment does not typically improve credit quality in the way a new revenue source or major economic development project might. More often, it protects against credit deterioration. That is still a powerful argument, but it is a different one. It says that resilience helps communities avoid worsening fiscal conditions, higher future costs, and greater pressure on households and public budgets.

That distinction matters because capital markets are generally better at rewarding visible gains than avoided harm. A new revenue source, a growing tax base, or a major economic development project can be easier to value because the financial benefit is direct. Resilience investment often works differently: it helps prevent worse outcomes, such as higher recovery costs, weaker revenues, disrupted services, or credit deterioration. That value is real, but it is harder to price because it often shows up as a loss that did not happen. There have also been very few climate-related municipal defaults, in part because FEMA and other federal safety nets have historically buffered state and local governments. The risk is tail risk—low-probability events that are nonetheless possible—and tail risk is systematically underpriced. Because that risk is difficult to price, it is often underrecognized until after a disruption occurs.

The clearest near-term opportunity is to turn this into a cost-of-delay argument that local governments and their finance partners can use. Paying for infrastructure before a disaster is usually cheaper than paying for recovery afterward. Because municipal bonds are often repaid through property taxes, delayed investment can mean higher future taxes, higher recovery costs, and more debt. This is an argument voters, budget offices, lenders, investors, and rating agencies can understand. It does not require solving the full challenge of pricing rare disasters in capital markets. It starts by showing what communities may pay if they wait.

Insurance offers a related but more developed pathway. Where credit markets are still learning how to price reduced physical risk, insurance has clearer evidence that verified risk reduction can change financial outcomes, even if insurance incentives cannot carry the full investment case alone.

Use Insurance as One Part of the Incentive Stack

Insurance offers one of the clearest near-term pathways for showing that verified risk reduction changes financial outcomes. IBHS data from Hurricane Sally showed a 70% reduction in insurance claim frequency and a 22% reduction in claim severity for FORTIFIED-designated properties. That is among the clearest existing evidence that verified risk reduction changes financial outcomes. Insurance is further along than most sectors in translating physical risk into pricing.

The opportunity is to use that evidence as a foundation for broader investment decisions, not to expect insurance incentives to carry the full return on investment alone. Insurance pricing is shaped by many forces beyond risk reduction, including inflation, labor and material costs, litigation, tariffs, interest rates, reinsurance costs, and market conditions. As one session participant put it directly: insurance discounts alone will not create a full return on investment for risk reduction.

That does not make insurance less important. It makes it a critical part of a larger incentive stack. Verified risk reduction should be able to inform insurance pricing, credit decisions, grants, tax policy, public investment, concessional capital, and regulation. The same evidence that helps an insurer understand lower claim frequency or severity could also help a lender understand lower credit risk, a public agency justify grant funding, or an owner make the case for investing in resilience.

For industrial facilities, where insurance can be one of the largest operating expenses, this connection could be especially powerful. If resilience technology adoption can be tied to demonstrable insurance outcomes, and if insurers reflect those outcomes in pricing or coverage terms, it could create a flywheel for broader adoption. If that connection does not materialize, facility owners may underinsure, and the state may become the insurer of last resort. The opportunity is to align incentives all the way through, so verified risk reduction is recognized not only by insurers, but by the full set of actors who benefit when losses are avoided.

These opportunities are promising, but they will not scale through isolated examples. They require a more deliberate set of handoffs among risk science, engineering, public budgets, capital markets, insurance, and communities.

What We Need to Build Next

This final section turns the workshop findings into an action agenda. It identifies the practical tools and handoffs needed to make resilience evidence usable across decisions: a metrics translation guide, shared evidence standards, real-world case testing, decision-ready dashboards, stronger finance arguments, long-term accountability, and trust with communities. The emphasis is not on creating one universal metric, but on building the conditions that allow risk reduction to shape financing, insurance, public investment, project delivery, and community outcomes.

Build Handoffs, Not a Universal Framework

A clear workshop takeaway was that the sectors working to measure, finance, insure, and deliver resilience do not need another universal framework or perfect metrics. They need a practical architecture for handoffs, so that evidence produced in one sector can inform decisions in another without being rebuilt from scratch every time.

Start with the Decisions Evidence Must Inform

That architecture has to work in the decision environments where resilience choices are actually made. In government, data matters, but political will, leadership vision, public trust, and voter support often determine whether bond measures, budgets, and policies move. An elected official may need a compelling story more than a probability chart. That is not a failure of rationality; it reflects how public-sector decisions are made under finite resources, competing priorities, and the need to build support for investments whose benefits may unfold over decades.

The same is true outside government. Companies most engaged in resilience often have leadership that has already decided it matters, not leadership that was persuaded by a model alone. Metrics need to support vision, not substitute for it. A better measurement system therefore has to speak not only to insurers, lenders, investors, and rating agencies, but also to elected officials, residents, community organizations, and business leaders.

Show Who Can Use Each Measure

A shared framework is useful only if it helps people use resilience measures in real decisions. It should not force every resilience benefit into a financial metric. Some benefits protect public budgets. Some protect households. Some affect insurance, credit, operations, or public trust. The point is to show what a measure tells us, who can use it, and what decision it can inform.

Table 3 turns that idea into a practical organizing tool. It is not a universal standard. It is a way to see where evidence needs to travel so that a measure produced in one setting can be useful in another.

Table 3. What different resilience measures can show

A cross-sector working group could build a metrics translation guide that identifies, for each measure, who already collects it, who else could use it, what decision it affects, what financial signal may change, and what societal value it protects.

This is a 12-month research and convening task. The crosswalk itself—not a universal standard, but a map of handoffs—is the specific missing piece that a decade of resilience framework-building has not yet produced.

Define What Counts as Proof

Before a resilience measure can change a decision, the person using it has to trust that it shows something real. Resilience investment depends on a prior belief that the investment works. IBHS standards and physical risk analytics platforms are mechanisms that create that trust for insurers and capital markets. But the evidence standards differ by sector, and no one has mapped them against each other. What do insurers need to believe a retrofit reduces risk? What do lenders need? What do communities need? Mapping those standards and where they diverge would accelerate adoption faster than any individual case study.

Test the Metrics Translation Guide on Real Cases

The metrics translation guide should be tested against live cases, not theoretical examples. DC stormwater credit trading, FORTIFIED homes in coastal Alabama, the Port Authority climate risk assessment, Montgomery County Green Bank resilience lending, The Nature Conservancy's nature-based resilience credits, and the heat/shade/commercial corridor question are all live examples ready for deeper translation work. Each one could produce a case study that shows where any proposed framework holds and where it breaks. These would be more useful than any case study designed to confirm the framework.

This testing should also define what trusted effectiveness actually requires. Resilience investment depends on a prior belief that the investment works. IBHS standards and physical risk analytics platforms are examples of mechanisms that create trust for insurers and capital markets. But evidence standards differ by sector. What do insurers need to believe a retrofit reduces risk? What do lenders need? What do communities need? Answering these questions will accelerate adoption faster than any individual case study.

Build Decision Tools for Different Audiences

A useful dashboard should distinguish investment indicators (what has been spent or built), implementation indicators (what has been completed), outcome indicators (what has changed), financial signals (what credit or insurance conditions have shifted), and societal impacts (who has been protected). A mayor's dashboard and an investor's dashboard may need to show different layers of the same underlying data. Transparency across all layers serves all audiences, but collapsing them into a single implementation count serves none.

Rather than treating resilience as a single concept, defining it through specific attributes—redundancy, robustness, flexibility, and adaptive capacity—makes it easier to track change over time and compare across projects and sectors. Formalizing the approach and connecting each attribute to the financial and societal signals it affects would make the field more interoperable without requiring agreement on a single definition.

Make the Finance Case Usable Before Disaster Strikes

The case for resilience needs to move from general risk reduction to finance-ready evidence that can support budgets, capital plans, and bond disclosures before disaster strikes. Communities that defer resilience and capital-maintenance investment may face higher long-term costs through increased asset deterioration, emergency repair needs, insurance and recovery costs, tax-base pressure, and, for exposed issuers, wider credit spreads or higher debt-service costs. These channels are increasingly visible to rating agencies and municipal investors and can translate into higher taxpayer or ratepayer burdens.

Public-sector and capital market methodologies also need better points of connection. Federal methodologies for defining taxpayer interest in infrastructure—how the Office of Management and Budget and other federal agencies justify spending on bridges, wastewater plants, and housing—are distinct from how capital markets prioritize investment. The gap forces local governments to build two entirely separate cases for the same project. Even modest commonality between these two frameworks would significantly reduce friction. Identifying the specific translation points where a shared definition would unlock cross-sector action is an achievable goal within reach of current institutions.

Keep Resilience Value Accountable over Time

Resilience value only lasts if someone remains accountable for performance after construction. A project funded for construction but not for ongoing operations may lose performance over time. Green infrastructure requires maintenance. Nature-based solutions must perform across decades. Capital plans must adjust as risk changes. At the point of investment, every project should be able to answer basic questions: Who owns the intervention? Who maintains it? Who pays recurring costs? What performance threshold triggers corrective action? Who monitors outcomes? Who is accountable if performance declines?

These questions determine whether resilience value lasts, and whether it remains credible to the insurers, lenders, and rating agencies that may have recognized it in the first place. Ongoing accountability is more than a reporting requirement; it is the condition under which the investment case holds.

Build Trust in Both Directions

Trust cannot run only toward insurers and capital markets. Any shared measurement framework will fail if it is not also credible to the communities most exposed to climate risk, including communities that have experienced resilience as a word used to justify disinvestment rather than support. Measurement that centers community outcomes, is transparent about who benefits and who does not, and connects to political accountability is measurement that can last.

"The goal is not a line connecting every sector to every other. It is a more interconnected set of handoffs—so that whatever evidence you produce in your sector, and however it impacts others, there is a clearer path to the decisions that matter. If we all move a little closer to that, the chain becomes easier to build."
—Mark Borsuk, Duke University, workshop closing remarks, April 22, 2026

Acknowledgments and More Information

The workshop at Duke in DC was organized by Duke RESILE and funded by Duke University's Pratt School of Engineering. It included experts from Duke's Pratt School of Engineering, Nicholas Institute for Energy, Environment & Sustainability, and Nicholas School of the Environment, and was aligned with the Duke Climate Commitment.

Authors and Affiliations

• Mark Borsuk, Ph.D., James L. and Elizabeth M. Vincent Professor of Civil and Environmental Engineering, Pratt School of Engineering, Duke University
• Victoria Salinas, Climate Leader in Residence, Duke University

Use of Artificial Intelligence Tools

Artificial intelligence tools, including Claude and ChatGPT, were used during the development of these proceedings to synthesize concepts, assist in drafting, and refine language and structure. Consistent with Duke University guidance on responsible AI use, AI-generated outputs were treated as provisional and subject to verification. All analysis, strategic direction, conclusions, recommendations, and final editorial decisions were led by the authors. This document reflects a human-led development and review process informed by practitioner insights shared during the convening, interdisciplinary collaboration, and responsible use of AI-enabled tools.

Citation

Borsuk, M, and V. Salinas. 2026. Making Resilience Count: Translating Risk Reduction into Investment Signals. Durham, NC: Nicholas Institute for Energy, Environment & Sustainability, Duke University. https://nicholasinstitute.duke.edu/publications/making-resilience-count-translating-risk-reduction-investment-signals