This memo is a part of the Community Power & Policy Partnership’s Housing Policy for Belonging, a project aimed at uplifting the great student work happening across UC Berkeley’s campus. We are so grateful to these authors for wrestling their 40 page reports, written for technical clients into 5 page summaries for a wider housing audience. These memos offer lessons for advocates, funders, legislators, and other change-makers and serve as a resource for future research.
True power lies in the collective courage to imagine and demand a better world, one where affordable housing, health, and climate resilience are not privileges reserved for the few, but fundamental rights accessible to all.
As climate change accelerates, so too does the urgency to ensure that all communities, particularly those historically underserved, are protected from its most dangerous impacts. Across California, government agencies are working to expand access to retrofit funding. Yet, fragmented eligibility criteria are often applied as passive checklists rather than as screening tools for active prioritization. This memo introduces a vulnerability-based targeting framework to support more precise, equitable allocation of retrofit resources, ensuring that investments reach those facing the greatest climate and health risks.
This research is based on a larger project initiated by the Health Equity, Policy and Planning (HEPP) Unit at the Alameda County Public Health Department (ACPHD). The methodological design and spatial analysis were developed with technical input from epidemiologists at ACPHD’s Community Assessment, Planning, and Evaluation (CAPE) team. I would like to thank ACPHD for inviting me to undertake this analysis, detailed in this StoryMap, and for supporting the dissemination of these findings to a wider audience.
Who gets to stay cool when the climate heats up?
As extreme heat events increase in frequency and intensity, the health and economic toll on vulnerable populations grows sharper. A series of home retrofit assistance programs has been implemented across various government departments to create healthier indoor environments, reduce energy expenditures, and advance building decarbonization. For example, California Department of Community Services and Development (CSD) is administering the California Weatherization Assistance Program (WAP) and distributing funding through Low-Income Home Energy Assistance Program (LIHEAP), funded by the Department of Energy (DOE). Current funding eligibility is determined by a household’s total monthly gross income falling below 200% of the Federal Poverty Level, as outlined in the DOE Weatherization Assistance Program (WAP) guidelines. Additional prioritization is given to vulnerable populations (by age/disability/energy burden) and to households located in Region 1 (Hot) in the state’s Climate Zone Map.
Figure 1: Building Climate Zones Established for Title 24 Energy Efficiency Standards
Source: California Energy Commission, last updated on February 8, 2024
State-funded programs like the Multi-Family (MF) Program of the California Low-Income Weatherization Program (LIWP), an initiative under California Climate Investments that leverages Cap-and-Trade proceeds to serve disadvantaged and low-income communities, also rely on income-based selection criteria in terms of household characteristics. Initially targeting communities identified as disadvantaged through CalEnviroScreen, the program has since expanded its service area to include low-income properties and households statewide, regardless of geographic designation. To qualify, affordable housing properties must have at least 66% of units occupied by households earning 80% or less of the Area Median Income (AMI), as defined by the U.S. Department of Housing and Urban Development (HUD).
California Energy Commission (CEC)’s Equitable Building Decarbonization Direct Install Program applies the same income thresholds for multifamily homes, plus if rent for at least 66% of units is affordable to such households. For single-family homes, eligible properties must be occupied by low-income households earning up to 80% of AMI. Additionally, the program restricts service to specific geographic areas, including: disadvantaged communities designated by SB 535, census tracts with median household incomes at or below 80% of the AMI, and areas identified as low-income by the Department of Housing and Community Development. Households located within one-half mile of a designated disadvantaged community are also eligible.
While CalEnviroScreen plays a critical role in identifying disadvantaged communities, effectively mapping exposure to environmental hazards among sensitive populations, it does not account specifically for heat-related environmental stressors, nor does it include key demographic and housing indicators such as disability status, homeownership, or the condition of occupied housing stock. This memo proposes the development of a geospatial targeting tool to address these gaps, combining income-based eligibility criteria with place-based prioritization to improve the precision and equity of home retrofit assistance programs. Guided by the principle of targeted universalism—advancing climate-safe housing for all by directing resources to those most at risk—this approach calls for funding to be allocated in a way that responds to the specific structural vulnerabilities of each community, and, where possible, each household.
Heat Vulnerability Index (HVI): a powerful tool in community climate resilience planning
The Heat Vulnerability Index is a common measure of population vulnerability to adverse health effects facing extreme heat events. Researchers have developed various methods for constructing Heat Vulnerability Indices and implemented diverse modeling approaches across jurisdictions, including Washington DC, New York State, Wisconsin, Philadelphia, as well as California municipalities such as San Francisco, Fresno, and San Diego.
Building off of the conceptual model of HVI: Vulnerability = Exposure + Sensitivity - Adaptive Capacity, we selected the following indicators based on the availability of geospatial datasets and feasible interventions:
- Heat Exposure: Extreme Heat Days, Impervious Surface Area, and Tree Canopy Coverage
- Heat Sensitivity: Population with a Disability, Population in Dependent Age Groups, Asthma Crude Prevalence Rate, and COPD Crude Prevalence Rate
- Heat Adaptive Capacity: Age of Renter-Occupied Housing Units, Rent-Burdened Households, Median Household Income, and Population Below 200% of the Federal Poverty Level
The indicators in this model not only account for residents’ housing-related health conditions but also capture the social determinants of health, particularly demographic and socioeconomic characteristics, as well as housing conditions. Innovatively, we used the age of the housing stock as a key proxy for housing quality. Negative weights were assigned to housing age per ACPHD’s specifications to reflect its strong inverse relationship with thermal performance. Given limited granular data on renter-specific incomes and rents in the American Community Survey (ACS) datasets, census tract-level median household income and poverty status collectively approximate household earnings, and rent burden data serve as proxy indicators for rental unit affordability.
Figure 2: Visualization of HVI Indicator Layers
Heat Sensitivity
Heat Adaptive Capacity
Given the scope of the original research, all data were filtered to Alameda County and analyzed at the census tract level. Eleven spatial data layers were aggregated using spatial summation. For detailed methodology, please see page 14 of the full report.
Figure 3: Spatial Distribution of HVI Scores and High-heat-vulnerable Census Tracts
Source: Author’s Analysis
Among the 96 census tracts classified as “high-heat-vulnerable” (HVI scores ≥31, as defined by ACPHD’s advocacy need), the majority are located in Oakland (54), followed by Unincorporated Areas (12), Hayward (9), San Leandro (8), Berkeley (7), Livermore (3), and one each in Pleasanton, Fremont, and a shared boundary between Unincorporated Areas and Hayward.
Collectively, these tracts account for just 25.6% of Alameda County’s total population, yet they exhibit disproportionately higher concentrations of people with disabilities (31.83%), children (26.93%), and individuals earning below 200% of the Federal Poverty Level (46.67%). Rent-burdened households make up 56.71% of total households in these areas—significantly higher than the countywide average of 47.74%. Although these tracts contain only 33% of the county’s rental units, they include 44% of renter-occupied housing units built before 1950. Elevated rates of chronic respiratory and cardiovascular disease prevalence are also persistently observed across this high-heat-vulnerable zone.
We also examined the geographic distribution of renters by race and ethnicity. Around half of all renters who identify as Some Other Race (54.25%), Black or African American (50.51%), or Hispanic or Latino (47.15%) live in these high-heat-vulnerable tracts. In contrast, only about one-fifth of White renters (19.7%) and one-fourth of Asian renters (24.0%) reside in these areas. With additional contextual information about local communities, outreach campaigns can be tailored to effectively disseminate funding opportunities and connect households with programs they both need and qualify for.
Figure 4: Who Rents in High-Heat-Vulnerable Census Tracts?
Source: Author’s Analysis
In addition to percentage data, we can also extract raw counts for each indicator to estimate the number of individuals with specific characteristics and the number of housing units in need of intervention—insights that can further inform the quantification of funding gaps for specific communities. At the census tract level, it is also possible to analyze the contribution of each vulnerability factor to the final HVI score, insights that can support advocacy for targeted resource allocation to address specific drivers of heat vulnerability.
The HVI modeling largely aligns with other tools for prioritizing climate investments, but including housing quality further highlights the urgent need in Oakland and Berkeley.
Figure 5: California Climate Investments Priority Populations Mapping Tool 4.0
Source: SB 535
Toward Smarter Tools and Stronger Outcomes
HVI modeling equips stakeholders with a targeting tool that produces quantitative evidence to inform and support advocacy efforts. Depending on resources and data availability, stakeholders might replicate this index in other jurisdictions, customize indicators based on their policy objectives and advocacy need, further include additional social determinants and environmental factors such as:
- Race
- Prevalence of other respiratory and cardiovascular diseases
- Health insurance coverage and proximity to medical facilities
- Language barriers, social isolation, and educational attainment
- Air quality
- Exposure to other environmental hazards or extreme events like wildfire
- Household air conditioning availability
- Access to public and private transit and proximity to green space.
The unweighted summation approach assumes equal predictive power across all indicators, potentially oversimplifying their relative contributions to heat vulnerability. More advanced weighting methodologies (like principal component analysis) could be developed to better model both individual indicators’ predictive ability and their synergistic effects.
Due to the aggregated nature of ACS estimates, this analysis is limited to tract-level approximations and cannot precisely identify individual renter households with intersecting risk factors and vulnerability contributors. Individual household-level screening is therefore necessary to bridge these data gaps—particularly to assess housing conditions not captured in existing datasets.
Monitoring and evaluation can be achieved by automating the HVI calculation process using scripted workflows in coding environments. By developing streamlined, replicable code, stakeholders can integrate updated datasets as they become available, allowing the index to be automatically recalculated at regular intervals. Periodic updates of the results can further support consistent tracking of progress and gaps over time.
Conclusion
As climate change accelerates, so does the urgency to ensure that no community is left behind in the fight for safe, healthy, and climate-resilient housing. This memo proposes the Heat Vulnerability Index (HVI) as an equity-focused targeting tool for home retrofit assistance programs—one that integrates environmental hazard exposures, social determinants of health, and housing-related risk factors.
Moving beyond current targeting mechanisms that separate income-based eligibility criteria and place-based prioritization, this tool is not meant to serve as a passive checklist to approve or reject applications. Instead, stakeholders could leverage existing datasets to pinpoint highly vulnerable communities, identify the specific drivers of risk, conduct targeted home visits, estimate funding gaps, and develop localized community outreach and policy advocacy strategies.
Investing in home retrofit programs guided by this tool is an investment in climate justice, health equity, and housing security. By intentionally directing resources to where the need is greatest, we can transform data into decisions—and decisions into durable, equitable outcomes.
Editor's note: The ideas expressed in this blog are not necessarily those of the Othering & Belonging Institute or UC Berkeley, but belong to the author.
