FAQs About Policies and NFIP Data

Non-sensitive data about the program is available in the Data and Analytics page.

The NFIP also regularly publishes anonymized policy-level and claims-level data through the OpenFEMA initiative.

Please note that due to the number of records and file size, it may need a data visualization or database querying tool. Instructions for interacting with the full dataset are included with each dataset. See: NFIP Redacted Claims and NFIP Redacted Policies.

NFIP policyholder data is now available through both Floodsmart and OpenFEMA for public entities including:

• Public institutions
• State and local governments
• Write Your Own (WYO) companies
• The NFIP’s Direct Servicing Agent (NFIP Direct)
• NFIP Vendors
• Academia
• Private insurers
• Capital markets
• Private sector industries
• Non-profit and foundations
• Other federal government agencies

Our program aims to ensure this data is refreshed every 40–60 days.

Under current federal law, people looking to buy a home can’t get information about past flood insurance claims on the property. Disclosure without the consent of the claimant and/or current homeowner is prohibited under the Privacy Act. However, some state laws may allow buyers to get certain flood-related information from non-federal sources.

Once someone becomes the new owner of the home, they are allowed to request details about flood insurance claims and payments made on the property—both during and before their ownership. This is allowed under FEMA’s System of Records Notice (SORN). The name of the requester, the type of record they are looking for and the required verification of identity should be included, when required by 6 CFR § 5.21.

Within FEMA, the NFIP is managed across several different constituent directorates. Information about community history or the CRS is maintained by the Floodplain Management Directorate (FMD). If you intend to request this information through the f of Information Act (FOIA) process or otherwise, please specify FMD as the location to be searched. The NFIP status of community in the declared counties is published in the NFIP Community Status Book.

FEMA / NFIP doesn’t make any direct determinations of substantial damage. That authority and responsibility rests with participating NFIP communities. Our program only receives information from participating communities when a claim receives payment following a “Substantial Damage” determination. This can include payments made under Improved Cost of Compliance (ICC) coverage. Details on ICC payments are included in the OpenFEMA redacted claims dataset.

FEMA believes this historic release of NFIP data promotes transparency, makes public data requests easier and helps people better understand and work with the program. This is the largest, most comprehensive release of NFIP data coordinated by FEMA to date. Academic institutions, the private sector and the public can access and conduct customizable searches, create reports, analyze, visualize and present historical NFIP data faster and easier than before.

OpenFEMA shows FEMA’s commitment to helping the nation prepare for floods by sharing important claims and policy information. This helps people make better informed decisions about their flood risk and the insurance they need to protect their homes and lives.

As part of the OpenFEMA initiative, FEMA is providing read-only, Application Programming Interface (API) based access to flood insurance policies and claims datasets. For information on how to customize queries using the API, please refer to the OpenFEMA Developer Resources page on the OpenFEMA website.

Anyone can use OpenFEMA to access the most current flood insurance claims and policies datasets (updated every 40–60 days). FEMA will post the program’s full claims history with a date range from 1978 to present, and a policy dataset with policy records since 2009.

Due to data retention policies in place prior to 2009, FEMA doesn’t have access to complete policy records from before this time. The data released by OpenFEMA represents the best accounting of NFIP policies for as long as the program has complete data. We can’t provide the same level of detail or volume for NFIP policies from the years 1970 to 2008 as we do in current OpenFEMA products. Any NFIP policy data from before 2009, if it was ever received or provided, should not be seen as a full or accurate picture of the policies in effect at that time. Most of the policy records from before 2009 that were kept are for policies that also had a flood insurance claim recorded, though not all. Please note, this only applies to NFIP policy records. All NFIP claims records are preserved.

NFIP policy data from before 2009, if ever received from or provided by our program, should not be considered representative of the policy base in force at that time. The policy records that were preserved from before 2009 are almost entirely, but not exclusively, policies for which a flood insurance claim was also recorded.

Please note that this only applies to NFIP policy records. All NFIP claims records are preserved.

This large dataset was segmented to improve accessibility and performance. The files have also been truncated and compressed. The first segment includes headers and metadata, while all additional segments don’t. The policy dataset is segmented based on record counts and not region or time frame.

Personal identifiable information (PII) is redacted. Data is also anonymized to the census tract, reported zip code and to one decimal point (.1) digit of latitude and longitude. If mapped, flood insurance policies and claims may appear to be clustered at a particular location due to this anonymization.

FEMA has a responsibility to protect policyholder privacy pursuant to the Privacy Act of 1974. In the June 2019 release, FEMA shared the most detailed data possible to promote transparency while protecting customer privacy consistent with the Privacy Act of 1974 and the Freedom of Information Act (FOIA). This is consistent with additional FEMA programmatic datasets posted on OpenFEMA as well.

NFIP policy data released on OpenFEMA is designed to provide insight into NFIP operations and flood risk in aggregate. It’s not designed to track individual structures or policy holders over time. Our program doesn’t provide a record ID which allows users to associate a specific claim with a specific policy record at this time. FEMA has a responsibility to protect policyholder privacy pursuant to the Privacy Act of 1974. FEMA has provided the most detailed data possible to promote transparency while protecting customer privacy consistent with the Privacy Act of 1974 and the Freedom of Information Act (FOIA).

FEMA is reporting the data to the public as it was reported and recorded in the program’s System of Record (SOR). The system of record includes business rules that reject critical transaction errors but may allow for some minor errors. This approach helps keep insurance available without the delays that would come from fixing every historical data issue. To reduce confusion and ensure quality data, FEMA has omitted fields in the current OpenFEMA dataset for which there is low confidence. In addition, the NFIP has some policy rules that allow properties to be rated in one flood zone, while physically located in another. Over time, the NFIP’s recorded and reported geographical data has become more precise. Therefore, more recent data will be more reliable for geographical analysis than older data.

FEMA doesn’t recommend using this data as input for commercial catastrophic models. The truncated coordinates don’t provide an accurate depiction of spatial risk and will not result in credible model results. In addition, using truncated coordinates may shift the location of risks, which can affect how accurate the elevation data is. FEMA recommends using the provided elevation data as a descriptive field for aggregation only. FEMA doesn’t intend to release assumptions for mapping OpenFEMA data to various commercial catastrophe model inputs.

Prior to the adoption of the NFIP’s new System of Record (SOR), PIVOT, in 2019, reporting requirements and practices were documented in Transaction Reporting and Process (TRRP) manuals. See all NFIP’s legacy TRRP manuals.

Since much of the data our program has released is historical in nature, many of these definitions and practices still apply. For example, where our system requires a value but isn’t always reported, “99XX” is often used as placeholder or default value.

We provide much of the data as it’s reported to our program by the NFIP Direct servicing agent and our private sector Write Your Own (WYO) partners. While the program’s PIVOT system has automated processes in place to reject or correct critical errors, it’s still possible for some incorrect data to be allowed into the system. When taken together, this means that abnormal data can at times persist in our program’s records. Therefore, it can be reflected in the data released through OpenFEMA and Floodsmart. If you believe you have identified an error within NFIP data, please contact us.

The policy file is a transactional dataset which shows when a policy was in force and eventually terminated. To understand which policies were active at a specific point in time, users should look for records where the “policyeffectivedate” is on or before the given date and the “policyterminationdate” is after that same date. Most NFIP policies have a standard policy period of one year but can be renewed indefinitely. Each subsequent renewal constitutes a separate record in this dataset.

The current claims data doesn’t include payment date or claims transaction. Users can use the date of loss to understand the magnitude of losses for a given calendar year, but it isn’t possible to report disbursements by year from this dataset.

Latitude and longitude can be used to identify a specific location. In order to protect policyholder privacy, latitude and longitude are truncated to one decimal point. The more significant digits (decimals), the more precise a set of coordinates can be. The level of data that FEMA provides for users will find the coordinates accurate to within approximately six miles.

Census tract and county fields are best used for aggregation. This is because they are derived from a policy or claim geocode (i.e. latitude and longitude are generated from provided address information). Therefore, we are confident that these values are reported with a relatively high degree of accuracy.

No, even under the FOIA process, certain release exemptions still apply. A more thorough description of the specific exemptions and exclusions for the Department of Homeland Security (DHS) are available at dhs.gov/foia-exemptions.

Decisions to withhold or redact information in FOIA requests are adjudicated on a case-by-case basis. In general, information can be withheld or redacted if releasing it would cause more harm to a person, protected group or national security than the benefits of release to the public.

When requesting NFIP data through the FOIA process, this means records with sensitive or confidential information may be redacted, unless the requester has obtained permission to release it.

Information that could harm a person’s privacy or a company’s competitive position is typically considered exempt from release under FOIA exemptions 6 and 4, respectively.

Due to privacy concerns, information that’s considered PII is normally withheld unless consent to release has been secured. This includes but isn’t limited to the following: (1) identity (e.g. name, policy #, repetitive loss ID, etc.), (2) address (e.g. street address, ZIP+4, etc.), and (3) location (e.g. census block, complete geocoordinates).

Please note that the address of an NFIP policyholder is considered PII by FEMA. Therefore, addresses of flood properties from our program can’t be shared publicly without violating the Privacy Act. This is in part related to our status as a federal agency with a database system of record that’s protected by the Privacy Act. Because of concerns about competitive advantage, information about policies issued or claims handled by NFIP’s private-sector partners is normally withheld. This includes but isn’t limited to the following information: WYO provider, insurance agency, insurance agent and claim adjusting firm. This isn’t meant to be an exhaustive list and has been shared here for information purposes only.

The NFIP doesn’t track flood insurance claims based on FEMA’s disaster declarations. The NFIP uses a separate Event Designation Number (EDN) or Flood Insurance Claims Office (FICO) number to do this. If an EDN was not assigned for a particular disaster, the NFIP can still search for flood insurance claims records using a state(s) and a date of loss range(s). Record level information may be subject to exemptions under the FOIA.

This depends on the nature and use of the dataset. Some datasets were used in the studies before the implementation of FEMA's current privacy initiatives and aren’t scheduled to be released via OpenFEMA or unredacted through the FOIA process. Other datasets may only have been available for academic use under authorized Information Sharing Access Agreements (ISAAs).

Some datasets were used in studies before FEMA's current privacy measures were put in place. These datasets aren’t planned for release through OpenFEMA or FOIA without redactions. Other datasets may have only been available for academic use through approved information sharing agreements (ISAAs).

Several factors go into rating an NFIP policy and calculating a premium. The total cost to consumers is inclusive of several congressionally mandated fees. These factors and calculations are laid out in the Flood Insurance Manual, which is updated semi-annually. Learn about the NFIP's pricing approach.