2023-24248. Agency Information Collection Activities; Submission for OMB Review; Comment Request; Federal Insurance Office Climate-Related Financial Risk Data Collection for U.S. Homeowners Multi-Peril Underwriting Data

AGENCY:

Federal Insurance Office, Departmental Offices, Department of the Treasury.

ACTION:

Notice of information collection; request for comment.

SUMMARY:

Pursuant to the Federal Insurance Office Act of 2010 (FIO Act), the Federal Insurance Office (FIO) of the U.S. Department of the Treasury will submit the following information collection request to the Office of Management and Budget (OMB) for review and clearance in accordance with the Paperwork Reduction Act of 1995 (PRA), on or after the date of publication of this notice. This data collection will assist FIO's assessment of the potential for any major disruptions of private insurance coverage in regions of the country particularly vulnerable to climate change impacts. The public is invited to submit comments on this request.

DATES:

Comments must be received by December 4, 2023 to be assured of consideration.

ADDRESSES:

Written comments and recommendations for the proposed information collection should be submitted at https://www.reginfo.gov/​public/​do/​PRAMain. Find this information collection by selecting “Currently under 30-day Review—Open for Public Comments” or by using the search function.

FOR FURTHER INFORMATION CONTACT:

Copies of the submissions may be obtained from Elizabeth Brown, Senior Insurance Regulatory Policy Analyst, by emailing [e-mail], calling (202) 597–2869, or viewing the entire information collection request at www.reginfo.gov.

SUPPLEMENTARY INFORMATION:

Title: Federal Insurance Office Climate-Related Financial Risk Data Collection for U.S. Homeowners Multi-Peril Underwriting Data.

OMB Control Number: 1505–NEW.

Type of Review: Request for a new OMB Control Number.

Description: On May 20, 2021, President Biden issued the Executive Order on Climate-Related Financial Risk, Exec. Order No. 14030 (E.O. 14030), which called on FIO to: (1) “assess climate-related issues or gaps in the supervision and regulation of insurers” and (2) “further assess, in consultation with States, the potential for major disruptions of private insurance coverage in regions of the country particularly vulnerable to climate change impacts.” ^[1] This information collection is necessary for FIO to advance its statutory mandates (including to monitor the extent to which traditionally underserved communities and consumers, minorities, and low- and moderate-income persons have access to affordable insurance products, and to monitor all aspects of the insurance industry) and to fulfill the second undertaking for FIO under E.O. 14030.^[2]

Under FIO's data collection, FIO will obtain consistent, granular, and comparable homeowners insurance data that is not otherwise publicly available on a nationwide level. This nationwide view is critical to understanding how climate-related financial risks impact families and individuals across state markets and the United States, and how these effects could impact the broader financial system.

FIO Authorities

Under the FIO Act, FIO's authorities include monitoring all aspects of the insurance sector, including identifying issues or gaps in the regulation of insurers that could contribute to a systemic crisis in the insurance sector or the U.S. financial system; monitoring the extent to which traditionally underserved communities and consumers, minorities, and low- and moderate-income persons have access to affordable insurance products; collecting data and information on and from the insurance industry and insurers; analyzing and disseminating data and information; and issuing reports regarding all lines of insurance that FIO monitors.^[3] Each of these authorities is relevant to FIO's planned data collection.

Efforts To Collect Data From Other Sources

Before FIO seeks to collect data directly from insurers, the FIO Act requires FIO to coordinate with relevant insurance regulators in the 50 states, the District of Columbia, and the five U.S. territories (collectively, the State Insurance Regulators), relevant federal agencies, and any publicly available sources in accordance with procedures set forth in the FIO Act before FIO collects the data directly from insurers.^[4] Prior to publishing FIO's October 2022 notice and request for comment related to the proposed data collection (October 2022 FRN), FIO coordinated with State Insurance Regulators to determine if they could provide in a timely manner the data that FIO proposed collecting.^[5] Following those efforts, as described in detail in the October 2022 FRN, FIO determined that the data that FIO was proposing to collect was not available or could not be obtained in a timely manner from any of the State Insurance Regulators, relevant federal agencies, or publicly available sources.^[6]

After the end of the comment period for the October 2022 FRN, FIO had additional meetings with the NAIC and the State Insurance Regulators to offer increased coordination and collaboration between FIO's proposed data collection and any efforts of the NAIC and State Insurance Regulators. On August 15, 2023, the NAIC announced a plan to issue a data call to assist State Insurance Regulators to collect data from insurers to better understand property markets and coverages and protection gaps, citing the “increasing frequency and severity of weather events, rising reinsurance costs, and inflationary pressures” that were creating insurance availability and affordability issues.^[7] FIO supports continued coordination and collaboration efforts that would enable it to fulfill its statutory mission while minimizing any burdens on insurers, to the extent feasible. It is unclear when the NAIC data call may be implemented and whether all State Insurance Regulators will participate in any proposed NAIC data call. FIO will continue to engage with the NAIC and monitor its development of a homeowners insurance data collection template and the use of that template or similar templates by State Insurance Regulators.

In sum, FIO has determined that the nationwide ZIP Code level data that it seeks to collect is not available or may not be obtained in a timely manner and has therefore determined to use its data-collection authorities under the FIO Act to obtain the necessary data directly from certain insurance groups.^[8]

Importance of Data Collection

The proposed data collection would provide FIO with a nationwide view of homeowners insurance that is critical to understanding how climate-related financial risks impact families and individuals across the United States and how these effects could impact the broader financial system. Since the publication of the October 2022 FRN, multiple homeowners insurers have announced decisions to exit certain areas or to decline to renew certain policies.^[9] As Treasury stated in a recent report: “Some evidence shows that states with exposure to climate hazards are already experiencing higher insurance costs, and the availability of insurance could have differential impacts for mortgage availability.” ^[10] Price increases often disproportionately impact traditionally underserved and disadvantaged communities and consumers, including those who are low- and moderate-income.^[11] Price increases could make insurance unavailable for such groups because they may be priced out of the market. A recent survey showed that about half of U.S. homeowners who did not purchase homeowners insurance had household incomes under $40,000 per year.^[12] FIO's work will facilitate further research and analysis to better understand the effects of climate-related risks on housing and the resulting ramifications for household finances and homeowners.

Purpose of Data Collection

FIO plans to use the information obtained from this data collection to perform a nationwide analysis to advance its statutory mandates and to respond to E.O. 14030. FIO will assess the impact of climate-related risk on the availability and cost of homeowners insurance in the United States. The primary goal of this data collection is to obtain consistent, comparable, and granular homeowners insurance data that is not otherwise publicly available that will enable FIO to assess the impact that climate-related physical risks have had on the homeowners insurance market since 2017 at a ZIP Code level. Data of the type FIO seeks is, at present, generally available only at the national or state level, with limited data available at a more granular level. Using the data from this collection, FIO will analyze nationwide trends, including comparisons of trends in homeowners insurance availability and costs. This data collection and analysis will benefit a wide variety of stakeholders, including insurers, reinsurers, homeowners, mortgage lenders, and policyholders, and inform policymakers by establishing a framework for granular, consistent, and comparable data with which to assess the localized impact of climate-related events and conditions on trends in homeowners insurance availability and cost. This data will help highlight underwriting trends and help identify areas where less-available or more-costly insurance is associated with higher risks from climate-related events. Such information may also help inform private and public efforts at the federal, state, and local levels to target pre-disaster mitigation measures and to improve policy in this area.

The narrower focus of this data collection includes data that will help establish a national baseline for analytics in this area. There is also potential value in collecting and analyzing more detailed information in this area. FIO intends to conduct climate-related data collections and analyses in the future, with a goal of doing so on an annual basis. In its initial analysis, FIO will focus on understanding nationwide trends in underwriting data. In future data collections, FIO may analyze in more detail, among other things, climate-related impacts on traditionally underserved communities or consumers, minorities, and/or low- and moderate-income persons.

FIO will not publish data or analysis of specific companies or groups or comparisons of results across companies or groups. FIO also will not publish data on or analysis of the solvency of insurance companies or groups based on the data collected. When FIO publishes analyses based on the data it collects, such analyses will be based on group-level submissions that will be aggregated across insurance groups to the ZIP Code level.

Data Collection's Key Elements

The October 2022 FRN provided the rationale for the proposed data collection and its eight key elements: (1) a focus on insurer underwriting, (2) insurance lines of business, (3) insurers, (4) data fields, (5) reporting framework, (6) reporting period, (7) geographic granularity, and (8) geographic scope.

Figure 1 below describes the initial proposed scope as well as FIO's changes to narrow that scope based in large part on stakeholder feedback. Among other things, FIO is removing data fields for which information is substantially similar to other fields, cannot be collected consistently at this time, or is not necessary for FIO's initial analysis. FIO believes that the narrower focus of this collection will improve data consistency and homogeneity and reduce the burden on the responding insurers.

Figure 1—Changes to FIO's Data Collection

Key elements	Proposed scope in October 2022 FRN	Summary of changes
Underwriting Focus	Physical risk from underwriting by P&C insurers, excluding liability risk 13	No changes.
Insurance Lines of Business	Homeowners Multi-Peril with 6 form types covering Mobile Homeowners, Owner Occupied Homeowners, and Other Homeowners policy form types	Homeowners Multi-Peril with 1 form type covering Owner Occupied Homeowners (form type HO–3, the most common homeowners policy form).
Insurers	Nationwide insurers writing above a premium threshold of $100 million in 2021 homeowners insurance direct premiums written Additional insurers in order to achieve at least an 80 percent market share threshold in each of 10 states that are potentially the most prone to climate-related disasters	Top nationwide homeowners insurance groups with 1 percent or more share of the homeowners insurance market based on direct premiums written in 2022, and request data to be aggregated and submitted at an insurance group level for all homeowners insurance entities identified within the group. FIO has identified 14 homeowners insurance groups with 240 homeowners insurance entities meeting the selection criterion.
Data Fields	15 data fields covering information regarding claims, premiums, and losses that correspond to data fields reported by U.S. insurers to State Insurance Regulators in annual filings, as well as additional policy information not collected on statutory filings	7 data fields (removing 8 data fields from proposed collection for which information is substantially similar to other fields, cannot be collected consistently at this time, or is not necessary for the proposed initial analysis).
		Revise 1 data field for clarity (to collect number of policies not renewed or retained rather than premiums not renewed).
		Remove exclusion of non-weather-related damages from reporting on claims and losses.
Reporting Framework	Accident year reporting basis	Accident year reporting basis for loss- and claims-related data, and calendar year reporting basis for premium-related data and policy information. (This change provides clarity on alignment with the format of statutory filings).
Reporting Period	5 years of underwriting data (2017–2021)	6 years of underwriting data (2017–2022).
Geographic Granularity	ZIP Code level for all U.S. ZIP Codes applicable to in-scope insurers	No changes.
Geographic Scope	Nationwide	No changes.

 

Underwriting Focus and Line of Business

FIO's data collection remains focused on the physical risk from underwriting by property & casualty insurers. The data collection also remains focused on the homeowners insurers multi-peril line of business. However, in large part in response to comments, this data collection will seek data relating to policies on a single form type (rather than the six forms proposed in the October 2022 FRN). Specifically, FIO will focus on owner-occupied homeowners multi-peril insurance policies issued on form “HO–3,” the most common homeowners insurance policy form in the United States.^[14]

Insurers

For this initial data collection, FIO plans to collect data from the top homeowners insurance groups that have 1 percent or more of the U.S. homeowners insurance market as measured by aggregate homeowners insurance direct premiums written in 2022. FIO has identified 14 homeowners insurance groups that meet this criterion (the Representative Sample Insurers). FIO's selection includes one standalone entity which is not part of a group. FIO's references to “group” include this standalone entity. These 14 groups (out of the 195 homeowners insurance groups operating in the United States) collectively underwrite about 70 percent of the direct homeowners insurance premiums written in the 50 states, the District of Columbia, and U.S. territories, based on 2022 financial statements. The Representative Sample Insurers have 240 homeowners insurance entities domiciled in 29 states and operating in many other states nationwide. FIO is taking a representative sample approach for two reasons: (1) FIO anticipates, for the purposes of its analyses, that the Representative Sample Insurers collectively have a sufficient market share for analyzing impacts in the U.S. market, and (2) the focus on the largest insurance groups will reduce the burden on small insurers.

Data Fields

FIO plans to collect data from the Representative Sample Insurers using the FIO Climate Data Collection: U.S. Homeowners Multi-Peril Form 2023–1 (Data Collection Form), which is an Excel spreadsheet that FIO will provide to them.^[15] In the October 2022 FRN, FIO proposed collecting 15 data fields. Based in large part on the comments that FIO received, FIO has decided to collect only seven data fields: (1) Number of Policy In-Force Exposures (on a calendar year basis (CY)); (2) Total Dollar Value of Coverage for Dwelling, Other Structures, Personal Property, and Loss of Use (CY); (3) Number of Policies Not Renewed or Retained (CY); (4) Direct Premiums Written (CY); (5) Total Direct Losses Paid (on an accident year basis (AY)); (6) Total Direct Incurred Losses (AY); and (7) Cumulative Number of Claims Closed With Loss Payment (direct claims on AY). Data from all seven fields is necessary for FIO to advance its objectives described above.

The Number of Policy In-Force Exposures field represents the count of policies that are in effect ( i.e., in-force) as of December 31 for each reporting calendar year, where one policy in-force corresponds to one home covered in that year. (This field does not include policies that lapsed or were canceled during the year). The field is necessary for FIO to calculate several metrics on a per-policy basis to allow for comparison across ZIP Codes. For example, FIO will calculate the average premiums, claims, losses, and policy non-renewals based on the total policies in a given ZIP Code; this field will also support calculation of changes in total policy counts over time.

The Total Dollar Value of Coverage for Dwelling, Other Structures, Personal Property, and Loss of Use field represents the total estimated value of insurance coverage for the dwellings or other property covered by the policies ( i.e., the potential exposures) in a ZIP Code, as recorded at the effective date of the policies or the date of their most recent renewal. This field is necessary because the premiums charged for policies will vary with the value of the properties insured as well as other factors. The dollar value of coverage and various calculated fields derived from that ( e.g., average premium per $1,000 in coverage) are commonly relied upon as metrics for accessing cost of coverage.^[16] The field will help FIO, among other things, assess the variation in the amount of financial protection for properties across ZIP Codes and over time for a given premium, while holding constant changes in insurance coverage. The field will also enable FIO to compare losses across ZIP Codes by calculating losses per $1,000 of insurance coverage.

The Number of Policies Not Renewed or Retained field addresses the number of policies that are not renewed or retained at any time during the reporting year, including non-renewals that are either insurer-initiated or policyholder-initiated. This field is necessary because understanding the percentage of policies not renewed will help FIO to assess availability issues. FIO notes that non-renewals alone may not be an adequate representation of availability issues because policies could be discontinued by the policyholder for various reasons, such as homeownership changes or switching to insurers that are not Representative Sample Insurers.

This field is replacing the field proposed in the October 2022 FRN that asked for the “Amount of Direct Premiums Written Renewed or Retained.” FIO is making this change because it is a more commonly reported measure of non-renewals. Some states ( e.g., California) and the NAIC's Market Conduct Annual Statement require insurers to report the number of policies not renewed or retained.^[17] To the best of FIO's knowledge, no state currently requires insurers to report “Amount of Direct Premiums Written Renewed or Retained.” Also, the “Number of Policies Not Renewed or Retained” is a subset of the data field that asks for the “Number of Policy in-force Exposures,” because in order to provide the “Number of Policy in-force Exposures,” an insurer would need to know both (1) the number of policies that it renewed or retained and (2) the number of policies not renewed or retained ( i.e., the information FIO is requesting).

FIO is retaining a data field for non-renewals because many commenters noted the importance of collecting information on non-renewals in order to understand where insurers may be pulling back coverage. Monitoring the number of non-renewals will allow FIO to understand non-renewal rates as distinct from changes in premiums.

The Direct Premiums Written field is a measure of policyholders' cost of obtaining coverage, and is necessary because the average premium per home or per $1,000 of coverage (premiums divided by number of policies or premiums divided by the amount of coverage) indicates what homeowners pay for property insurance on a comparable per-unit basis. That information, in turn, is necessary because changes in premiums over time and comparisons of average premium per home on a nationwide basis can provide meaningful insights into insurance costs. Collecting this data on a calendar year basis will provide FIO with up-to-date information on pricing trends. Premium data can also support analysis of the availability of homeowners insurance in ZIP codes.

The Total Direct Losses Paid field is necessary to understand how loss trends in each ZIP Code affect the availability and cost of homeowners insurance. Direct losses paid is the amount of money to date (for each accident year) that has been paid to, or on behalf of, policyholders, and does not include estimates of losses that have been incurred but not yet paid. FIO will analyze trends in paid losses (including on a per claim basis) by ZIP Code. FIO will also use this field to examine trends by ZIP Code in paid loss ratios (by dividing by premiums) and claims severity (by dividing by claims closed with loss payment) to better understand how underwriting metrics may reflect weather-related events.

The Total Direct Incurred Losses field is necessary because it provides a comprehensive measure of the value of an insurer's loss experience in a reporting period by ZIP Code. Unlike losses paid, incurred losses reflect the total value of losses that an insurer incurs during a given period, regardless of whether the losses have been paid out ( i.e., it includes reserves), and is used to estimate total losses for a given year. FIO will analyze trends in incurred losses (including on a per-claim basis) by ZIP Code. FIO also will use this field to look at trends by ZIP Code in incurred-loss ratios (by dividing by premiums) to better understand how underwriting metrics may reflect weather-related events.

The Cumulative Number of Claims Closed With Loss Payment field is necessary because the average amount insurers pay for each claim will provide insight on the impact of an event (or events) in a given area, helping FIO to identify exposure frequency and severity.^[18] During the underwriting process, an insurer must gauge the average value it will pay out per claim.^[19] Historically, the average cost of claims has been one of the primary factors behind the price increases and the tightened availability of homeowners insurance.^[20] FIO will be evaluating claims frequency and severity on a nationwide basis to see if they are associated with average premiums growth or a decrease in the availability of homeowners insurance. This data field will thus enable FIO to better understand how underwriting metrics may reflect weather-related events.

The seven data fields described above include three data fields not collected by State Insurance Regulators (the number of policy in-force exposures; the number of policies not renewed or retained; and the total dollar value of coverage for dwelling, other structures, personal property, and loss of use). The seven data fields also include four fields to collect ZIP Code level information nationwide regarding premiums, claims, and losses that correspond to four state-level data fields reported by U.S. insurers to State Insurance Regulators in annual filings.

Reporting Framework

The Data Collection Form and accompanying instructions now clarify that Representative Sample Insurers should use accident year reporting basis for loss- and claims-related data and calendar year reporting basis for premium-related data and policy information.^[21] This change to the instructions provides clarity on alignment with the format of statutory filings; there is no fundamental change.

Reporting Period

FIO seeks to collect six years of underwriting data, 2017 through 2022 (Reporting Period). This is a one-year expansion over the Reporting Period proposed in the October 2022 FRN. Expanding the reporting timeframe by one additional year will allow FIO to capture more current data, including the impact of recent climate-related disasters such as Hurricanes Ian and Nicole, and responds to stakeholder feedback recommending a longer Reporting Period.

Geographic Scope and Granularity

FIO will collect data at a ZIP Code level for all U.S. ZIP Codes in which the homeowners insurance entities within the Representative Sample Insurers have written owner-occupied multi-peril homeowners multi-peril policies corresponding to policy form HO–3 during the Reporting Period in order to conduct a granular, nationwide assessment to advance its statutory mandates and respond to the second undertaking described in E.O. 14030. The October 2022 FRN outlines several reasons why ZIP Code level data is critical for FIO's analysis, and those reasons remain valid.^[22]

Collection Process

Data will be collected from the Representative Sample Insurers on the Data Collection Form, which is a revised version of the template discussed in the October 2022 FRN. A copy of the Data Collection Form and the instructions for filling it out can be found at https://home.treasury.gov/​policy-issues/​financial-markets-financial-institutions-and-fiscal-service/​federal-insurance-office/​reports-notices. Each of the Representative Sample Insurers will be asked to aggregate and report the nationwide data requested by the Data Collection Form at a ZIP Code level for all owner-occupied homeowners multi-peril policies corresponding to policy form HO–3 that the homeowners insurance entities within its group have written during the Reporting Period.

Changes in Data Collection Based on Comments and Feedback Received

In response to the October 2022 FRN, FIO received 35 formal written comments with over 9,400 signatories. Commenters included individuals, brokers, insurance industry trade associations, State Insurance Regulators, public interest groups, consumer advocates, climate and environmental groups, and others. Following the close of the comment period of the October 2022 FRN, FIO met with a wide range of stakeholders, including the NAIC and some State Insurance Regulators, large writers of homeowners multi-peril insurance, and consumer and environmental groups, regarding the proposed data collection. In addition, FIO and members of the Federal Advisory Committee on Insurance (FACI) discussed the proposed data collection during two public FACI meetings and at meetings of the FACI Climate Subcommittee.

Based on the public comments and stakeholder feedback, FIO has revised several aspects of its data collection, including the scope of the forms included in the homeowners insurance multi-peril line of business, the selection of insurers, the number of data fields, and the reporting period for the collection. FIO also revised the instructions for the Data Collection Form to clarify how the data should be reported, such as indicating that premium-related data and policy information should be provided on a calendar year basis and requiring all claims and losses to be reported, not just weather-related claims and losses. These changes, taken together, should lessen the burden on the insurers while also enhancing FIO's analyses.

Revised Focus Within Homeowners Line of Business

FIO has revised its data collection to focus only on the owner-occupied homeowners multi-peril policy form known as HO–3. In response to the October 2022 FRN, insurers highlighted the difficulty of collecting consistent, comparable data across multiple policy form types. Limiting the collection to one form will allow FIO to collect more consistent, comparable data as compared to the proposal set forth in the October 2022 FRN, which proposed collecting six homeowners insurance policy forms.

Revised Data Fields

Based on the comments received and FIO's decision to narrow the scope of the data collection, FIO removed eight of the fifteen data fields proposed in the October 2022 FRN. FIO removed fields for which the information is substantially similar to other fields ( e.g., direct premiums earned), cannot be collected consistently at this time ( e.g., total dollar amount of insurance deductible), or is not necessary for FIO's initial analyses ( e.g., number of claims reported). The narrower focus of this data collection includes data that will help establish a national baseline for analytics in this area.

Revised Insurer Selection

FIO did not receive many comments related to the proposed methodology for selecting the insurers for the proposed data collection, although it did receive comments on the overall burden of the proposed data collection, as well as on the potential use of the data for entity-specific analyses, and the operational burden of having groups report their subsidiaries on separate templates. In the October 2022 FRN, FIO's proposed selection criteria for insurers were: (1) nationwide insurers writing above a premium threshold of $100 million in 2021 homeowners insurance premiums; and (2) additional insurers in order to achieve at least an 80 percent market share threshold in each of 10 states that are potentially the most prone to climate-related disasters.

Although FIO did not receive many specific comments on the selection of insurers, FIO is changing the selection criteria to cover only the largest homeowners insurance groups, i.e., those with 1 percent or more of the U.S. homeowners market nationwide, resulting in the selection of the top 14 homeowners insurance groups by direct premiums written in 2022. FIO has revised the selection criteria in order to: (1) reduce the burden on smaller insurers; (2) alleviate concerns that FIO would be conducting entity-specific analyses; and (3) make the data operationally easier for insurers to report and for FIO to collect and analyze.

Revised Reporting Period

FIO is increasing the reporting period from the proposed five years (2017–2021) to six years (2017–2022) because expanding the reporting timeframe by one additional year will allow FIO to capture the impact of recent climate-related disasters, including Hurricanes Ian and Nicole. A number of commenters supported a longer reporting period, although many of these called for adding multiple additional years. The overall burden for the Representative Sample Insurers will not increase, because FIO is removing eight of the fifteen data fields that it proposed in the October 2022 FRN and focusing on only one policy form (HO–3), instead of six policy forms.

No Change to Geographic Scope or Granularity

FIO received many comments on the proposal to collect data at the ZIP Code level, both for and against.^[23] FIO plans to move forward with ZIP Code level collection. As many commenters noted, state-level data would be insufficient for FIO's analysis because climate-related risks and socioeconomic factors vary across geography, and state-level trends could differ from trends seen on a more granular level. While FIO understands the potential additional benefits of collecting census tract level data relative to ZIP Code information, FIO's discussions with insurers and other stakeholders leads FIO to conclude that ZIP Code information tends to be collected by insurers as part of the address in a homeowners policy, while other geographic information, such as census tracts, may not be collected by insurers. Consequently, collecting data at a census tract level could increase the burden on insurers to produce information at that granularity. FIO believes that ZIP Code collection most appropriately balances the need for granular data with the responsibility to minimize collection burden.

Revised Instructions

FIO has revised the instructions for the Data Collection Form relating to the data fields. The proposed template instructions provided with the October 2022 FRN directed insurers to report data only for losses and claims for weather-related damages. In discussions with FIO, however, insurers noted that it can be difficult or impossible to exclude non-weather-related damage from other loss estimates, based on how information is stored within insurers' systems. In addition, information from the Insurance Information Institute shows that about 90 percent of homeowners insurance losses have been due to physical damage from wind, hail, water damage, freezing, fire, and lightning.^[24] Therefore, FIO has modified the instructions for the Data Collection Form to remove the exclusion of non-weather-related damages in connection with the reporting of claims and losses.

Revised Burden Estimate

FIO received many comments, including in follow-up stakeholder engagements, regarding its burden estimates, including the number of hours estimated (hour burden) and average salary used (hourly rate). Multiple comments stated that the hour burden estimate was too low and that the hourly rate should be higher to reflect the need for subject matter experts in actuarial and finance roles to respond to the collection. Other commenters supported FIO's estimate and emphasized that the importance of the information collected outweighed the burden.

FIO has revised its estimate of the hour burden to reflect that the data will be collected for: (1) only seven data fields, not the 15 data fields originally proposed; (2) only one policy form of homeowners insurance, not six as originally proposed; (3) one additional year, 2022, beyond the original proposed five years (2017–2021); and (4) the change in insurer selection criteria.^[25]

FIO updated the estimated hourly rate to reflect the salary and benefit increases between June 2022 and July 2023 reported by the U.S. Bureau of Labor Statistics. FIO has not, however, made any other adjustments to the hourly rate because the FIO data collection is largely an operational task focused on data already collected by insurers.

FIO estimates the annual burden for its data collection is between 14,400 and 48,000 hours at a total cost of $839,520 to $2,798,400 for the Representative Sample Insurers, collectively.^[26] For reference, the Representative Sample Insurers reported aggregate revenues of over $372 billion in 2022.^[27] FIO's revisions to its data collection have lowered the burden estimate by about 30 per as compared to the estimates in the October 2022 FRN (Figure 2). In addition, the Representative Sample Insurers are the 14 largest homeowners insurance groups and, as a result, they may experience internal synergies and efficiencies when completing the Data Collection Form. Thus, the total number of hours that it may take all Representative Sample Insurers to collect, process, and complete the Data Collection Form may be less than the number of hours that FIO has estimated here.

Figure 2—Comparison of Burden Estimates

	October 2022 FRN	Current estimates
Hours per homeowners insurance entity	100 to 350	60 to 200
Total hours for all Representative Sample Insurers	21,300 to 74,550	14,400 to 48,000
Hourly rate	$54.27	$58.30
Total cost for all Representative Sample Insurers	$1,155,951 to $4,045,829	$839,520 to $2,798,400

Analysis of Data Collected

FIO currently plans to analyze the data collected using trend or time-series analysis. In the time-series analysis, FIO will assess trends in underwriting, claims, and loss metrics, normalized on a per-policy, premium, or coverage basis to allow for comparison across ZIP Codes, for owner-occupied homeowners multi-peril (HO–3 form) policies over the last six years. In addition to the data collected on the Data Collection Forms, FIO's analysis will rely on several types of data that may include, but are not limited to, publicly available insurance information such as residual market information, information on the relative risks that locations face from climate-related events, and data used to control for other changes in a ZIP Code that could influence insurance markets ( e.g., inflation, real estate values, or changes in population). At a later stage, FIO may consider using additional methodologies, including multivariate statistical analysis.

Submission of Data

Reporting under this data collection would be mandatory for all Representative Sample Insurers. If OMB approves this data collection, the Representative Sample Insurers will be expected to submit the completed Data Collection Form through a secure Treasury web portal within 90 days of FIO publishing a Federal Register notice announcing the start of the data collection. (FIO has extended the time period which the Representative Sample Insurers will have to submit the information from 60 days to 90 days to ensure that they have sufficient time to gather and provide the requested data.)

FIO intends to provide training and additional resources within the data collection period to facilitate the proper completion of reporting templates.

Given the potential sensitivity of some of the requested data, FIO will seek to maintain the data submitted in a confidential manner. The FIO Act includes provisions addressing the privacy or confidentiality of submissions of non-publicly available data and information to FIO.^[28] In accordance with the FIO Act, submissions pursuant to this data collection will not constitute a waiver of, or otherwise affect, any privilege arising under federal or state law to which the data or information is otherwise subject.^[29] FIO expects that data it receives under this collection will likely contain or consist of “trade secrets and commercial or financial information obtained from a person and privileged or confidential.” This type of information is subject to withholding under exemption 4 of the Freedom of Information Act.^[30]

All data collection is expected to be completed through a secure portal maintained by Treasury, and FIO will not publish confidential firm-specific data from individual submissions. FIO may publish aggregated analyses of the submitted information.

Form: FIO Climate Data Collection: U.S. Homeowners Multi-Peril Form 2023–01.

Affected Public: Businesses and other for-profit institutions and not-for-profit entities.

Estimated Number of Respondents: 14.

Frequency of Response: On occasion, although FIO aims to conduct annual data collections.

Estimated Total Number of Annual Responses: 14.

Estimated Time per Response: Varies from 60 hours to 8,800 hours depending upon the number of homeowners insurance entities within a respondent group. The estimated total annual burden hours are 60 to 200 hours per homeowners insurance entity within a respondent group. The number of homeowners insurance entities within a respondent group varies from 1 to 44, with the average number being 17. Thus, the average estimated time per response ranges from 1,020 hours to 3,420 hours.

Estimated Total Annual Burden Hours: Approximately 14,400 to 48,000 hours total for all respondents together. The estimated total annual number of burden hours for each respondent will be based on the number of homeowners insurance entities within each respondent group. The estimated total annual burden hours are 60 to 200 hours per homeowners insurance entity within a respondent group.

Authority:44 U.S.C. 3501 et seq.

Spencer W. Clark,

Treasury PRA Clearance Officer.

Footnotes

1.  Executive Order on Climate-Related Financial Risk, Exec. Order No. 14030, 86 FR 27,967 (May 20, 2021) (E.O. 14030).

2.  31 U.S.C. 313(c); E.O. 14030. FIO addressed the first undertaking for FIO under E.O. 14030 with the publication of its report Insurance Supervision and Regulation of Climate-Related Risks (June 2023), https://home.treasury.gov/​system/​files/​311/​FIO-June-2023-Insurance-Supervision-and-Regulation-of-Climate-Related-Risks.pdf.

3.  31 U.S.C. 313–314.

4.  31 U.S.C. 313(e)(4) provides that “Before collecting any data or information under paragraph (2) from an insurer, or affiliate of an insurer, the Office shall coordinate with each relevant Federal agency and State insurance regulator (or other relevant Federal or State regulatory agency, if any, in the case of an affiliate of an insurer) and any publicly available sources to determine if the information to be collected is available from, and may be obtained in a timely manner by, such Federal agency or State insurance regulator, individually or collectively, other regulatory agency, or publicly available sources. If the Director determines that such data or information is available, and may be obtained in a timely manner, from such an agency, regulator, regulatory agency, or source, the Director shall obtain the data or information from such agency, regulator, regulatory agency, or source. If the Director determines that such data or information is not so available, the Director may collect such data or information from an insurer (or affiliate) only if the Director complies with the requirements of subchapter I of chapter 35 of title 44, United States Code (relating to Federal information policy; commonly known as the Paperwork Reduction Act), in collecting such data or information. Notwithstanding any other provision of law, each such relevant Federal agency and State insurance regulator or other Federal or State regulatory agency is authorized to provide to the Office such data or information.”

5.  Agency Information Collection Activities; Proposed Collection, Comment Request; Federal Insurance Office Climate-Related Financial Risk Data Collection, 87 FR 64,134 (October 21, 2022) (October 2022 FRN).

6.  October 2022 FRN, 87 FR at 64,140.

7.  NAIC, “NAIC to Issue Data Call to Help Regulators Better Understand Property Markets,” news release, August 15, 2023, https://content.naic.org/​article/​naic-issue-data-call-help-regulators-better-understand-property-markets.

8.  31 U.S.C. 313(e).

9.   See, e.g., Christopher Flavelle, et al. “Climate Shocks Are Making Parts of America Uninsurable. It Just Got Worse,” New York Times May 31, 2023, https://www.nytimes.com/​2023/​05/​31/​climate/​climate-change-insurance-wildfires-california.html.

10.  U.S. Department of the Treasury, The Impact of Climate Change on American Household Finances (September 2023), 12, https://home.treasury.gov/​system/​files/​136/​Climate_​Change_​Household_​Finances.pdf.

11.   See, e.g., White House Council of Economic Advisors, Economic Report of the President (March 2023), Box 9–2, 281, https://whitehouse.gov/​wp-content/​uploads/​2023/​03/​ERP-2023.pdf; U.S. Department of the Treasury, The Impact of Climate Change on American Household Finances, 6.

12.  Insurance Information Institute, Homeowners Perception of Weather Risks: 2023 Q2 Consumer Survey (2023), 2, https://www.iii.org/​sites/​default/​files/​docs/​pdf/​2023_​q2_​ho_​perception_​of_​weather_​risks.pdf.

13.  Liability risk is “a risk where an insured is liable to a third party as a result of or caused by any act, error, omission representation or statement by the insured.” “Liability Risk Definition,” Law Insider, https://lawinsider.com/​dictionary/​liability-risk.

14.  HO–3 policies comprise 55.5 percent of all homeowners insurance policies and 78.3 percent of all owner-occupied homeowners insurance policies. See NAIC, Dwelling Fire, Homeowners Owner-Occupied, and Homeowners Tenant and Condominimum/Coorperative Unit Owner's Insurance Report: Data for 2020 (December 2022), 4, https://naic.soutronglobal.net/​Portal/​Public/​en-GB/​DownloadImageFile.ashx?​objectID=​9803&​ownerType=​0&​ownerId=​2006.

15.  For more details on the data fields, a copy of the Data Collection Form and the instructions for filling it out can be found on FIO's webpage, https://home.treasury.gov/​policy-issues/​financial-markets-financial-institutions-and-fiscal-service/​federal-insurance-office/​reports-notices.

16.   See, e.g., NAIC, Dwelling Fire, Homeowners Owner-Occupied, and Homeowners Tenant and Condominium/Cooperative Unit Owner's Insurance Report: Data for 2020 (2022), Table 3, 4, and 5, https://naic.soutronglobal.net/​Portal/​Public/​en-GB/​DownoadImageFile.ashx?​objectId=​9803&​ownerId=​2006;​ Massachusetts Division of Insurance, Annual Home Insurance Report for Calendar Year 2020, 11, https://www.mass.gov/​doc/​the-2020-massachusetts-market-for-home-insurance/​download.

17.   See California Department of Insurance, “FACT SHEET: Data on Insurance Non-Renewals and FAIR Plan,” news release, December 20, 2021, https://www.insurance.ca.gov/​0400-news/​0100-press-releases/​2021/​upload/​nr117DataNon-RenewalsandFAIRPlan12202021.pdf; NAIC, Property & Casualty Market Conduct Annual Statement Homeowner Data Call & Definitions, https://content.naic.org/​sites/​default/​files/​inline-files/​industry_​mcas_​data_​call_​def_​homeowners_​2020.pdf. While the “Number of Policies Not Renewed or Retained” is not an annual statutory filing field, the NAIC collects statewide information on “Number of Company-Initiated non-renewals during the period,” “Number of cancellations for non-pay or non-sufficient funds,” and “Number of cancellations at the insured's request” from insurers writing homeowners policies as part of its Market Conduct Annual Statement data call, and these fields collectively sum to the number of policies not renewed or retained.

18. LexisNexis Risk Solutions, “Annual U.S. Home Insurance Trends Report Confirms Upward Trend in Loss Cost and Severity across All Perils,” news release, October 13, 2022, https://risk.lexisnexis.com/​about-us/​press-room/​press-release/​20221013-home-insurance-trends-report; Pat Howard & Kara McGinley,” Homeowners Insurance Claims Statistics in 2023,” Policygenius, December 22, 2022, https://www.policygenius.com/​homeowners-insurance/​homeowners-insurance-statistics/​.

19. AM Best, Best's Guide to Understanding the Insurance Industry (2023), 6, https://web.ambest.com/​information-services/​sales-information/​best-s-guide-to-understanding-the-insurance-industry.

20. See Martin Grace & Robert Klein, “Homeowners Insurance: “Markets Trends, Issues and Problems,” SSRN Electronic Journal, August 2003, https://www.researchgate.net/​publication/​228227047_​Homeowners_​Insurance_​Market_​Trends_​Issues_​and_​Problems.

21. Insurance information may be reported on either a calendar year or an accident year basis, depending on the type of information being reported. Accident year reporting monitors loss development of a claim over time (and typically includes losses when they occur, not when they are reported). For example, if a hurricane occurs in 2017, then 2017 would be the accident year. Losses and claims associated with the hurricane could occur in 2017 but could also develop into subsequent years as policyholders submit claims and receive payments over time but would be reported for the 2017 accident year. On the other hand, calendar year reporting does not change or develop over time beyond the calendar year. For example, premiums based on a 2017 calendar year would include premiums from January 1 to December 31, 2017.

22. See, e.g., October 2022 FRN, 87 FR at 64,138.

23.  Commenters that opposed ZIP Code level collection cited, among other things, the burden to insurers, the potential for misleading conclusions based on ZIP Code level claims data, and the possibility of exposing personally identifiable information or proprietary data in ZIP Codes with few policies. Those in favor of granular data collection tended to encourage FIO to collect data at an even finer granularity than ZIP Code, with census tract being the most common unit proposed. Those that preferred census tracts cited various rationales, including that they are comparable, have similar numbers of residents, tend to be internally similar, and align with census-defined geographies better than ZIP Codes.

24.  “Facts + Statistics: Homeowners and Renters Insurance,” Insurance Information Institute, https://www.iii.org/​fact-statistic/​facts-statistics-homeowners-and-renters-insurance.

25.  The Representative Sample Insurers encompass 240 homeowners insurance entities, which is more than the 213 entities that would have been covered under the selection criteria outlined in the October 2022 FRN. Any increase in burden from including more entities will be offset for two reasons. First, FIO is requesting that the 14 Representative Sample Insurers aggregate responses across all of their insurance entities, which should reduce the burden on any given entity and allow the groups to submit one consolidated response. Second, FIO is reducing the number of data fields and number of policy forms for which data should be collected, which will enable reduction of burden relative to the proposed estimate, even with the proposed change to add one more year of reporting.

26.  FIO estimates that each individual homeowners insurance entity within a Representative Sample Insurer will take between 60 and 200 hours total to collect the requested data for six years (2017 to 2022). The 14 Representative Sample Insurers have 240 homeowners insurance entities (with the number of entities per group ranging from 1 to 44, so the burden for each Representative Sample Insurer would vary). The overall estimated burden hours are 14,400 (240 entities × 60 hours) to 48,000 (240 entities × 200 hours). The average fully loaded hourly rate for insurance employees in July 2023 was $58.30. Based on data from “Insurance Carriers and Related Activities: NAICS 524,” U.S. Bureau of Labor Statistics, https://www.bls.gov/​iag/​tgs/​iag524.htm, the average wage rate for all insurance employees was $43.31 in July 2023, and the total benefit compensation in the 2nd Quarter of 2023 was 34.6 percent, which is a benefit multiplier of 1.346. Therefore, a fully-loaded wage rate for insurance employees is $58.30, or $43.31 × 1.346. Thus, the total cost for all Representative Sample Insurers to comply with the data collection would be between approximately $839,520 (14,400 hours × $58.30/hour) and $2,798,400 (48,000 hours × $58.30/hour).

27.  Data on revenues for the 13 groups was derived from S&P Global Market Intelligence (S&P Global) on October 18, 2023. Data for the revenue of the standalone entity came from its annual statement for 2022.

28.  31 U.S.C. 313(e)(5).

29.  31 U.S.C. 313(e)(5)(A).

30.  5 U.S.C. 552(b)(4).