Data and methods

Everything this site says about an insurer, a city or a ZIP code comes from public records. This page lists those records, explains each calculation we make with them, and sets out the limits of the data. The numbers on the site were last refreshed on November 22, 2026.

The datasets behind this site

DatasetUsed forLast updated by TDIOpen the data
Complaint indexes and policy counts for insurance companiesConfirmed complaints, policies in force and complaint index for each companyNovember 21, 2026data.texas.gov/d/pa9u-9s9w
Residential and Farm Policies in Force by ZIP CodeHomeowners, dwelling and renters policies in force by ZIP codeSeptember 3, 2026data.texas.gov/d/i22c-h2mm
Residential and Farm Policies in Force by CountyPolicies in force by countySeptember 3, 2026data.texas.gov/d/8mvr-4gj9
89R HB 2067 – actual number of residential property declinations, cancellations, and nonrenewalsNon-renewals, cancellations and declined applications by ZIP code and monthOctober 23, 2026data.texas.gov/d/m7yx-zxf2
89R HB 2067 – reasons for residential property declinations, cancellations, and nonrenewalsReasons insurers gave for those actionsOctober 23, 2026data.texas.gov/d/3efz-d6qn
Home and auto rate filingsRate filings by companyNovember 15, 2026data.texas.gov/d/iubg-btfs

All six datasets are published by the Texas Department of Insurance (TDI) on the State of Texas open data portal and can be downloaded by anyone without an account. We also use two U.S. Census Bureau files from the 2020 Census that describe how ZIP Code Tabulation Areas overlap cities and counties.

A few statewide figures come from TDI web pages that have no data feed: the average premium by year and the market share of insurer groups, from TDI's homeowners insurance market overview, and each company's share of premium, from TDI's top 40 list of insurers. We copy those by hand and record the date we read them.

Complaint index

TDI describes the index this way: a company's percentage of confirmed complaints for a line of insurance, divided by its percentage of policies in force for that line. The average is 1.00. Below 1 means fewer complaints than average; above 1 means more.

  • The line is broader than "homeowners". TDI's Homeowners line counts homeowners, renters, condo, dwelling and farm policies together. A company that mostly sells renters insurance is measured on the same scale as one that mostly insures houses.
  • Company figures are TDI's. For a single legal company we show TDI's complaint count, policy count and index unchanged.
  • Brand figures are ours. Brands such as Allstate or USAA write policies through several legal companies. On a brand page we add up those companies' complaints and policies and apply TDI's formula to the totals. TDI does not publish that combined figure.
  • How we group companies. A company joins a brand only when its legal name carries the brand or the affiliation is a matter of public record, for example Safeco under Liberty Mutual. Any other company with at least 20,000 policies gets a page under its own legal name. Corporate ownership changes, and we may have missed an affiliate; the table on each brand page shows exactly which companies were counted.
  • Small numbers move a lot. For a company with a few thousand policies, a single complaint can double the index. We flag any index built on fewer than 10 complaints and always show two years.
  • The labels. We describe an index above 1.20 as "more than average", below 0.80 as "fewer than average", and anything between as "near average". Those cut-offs are ours, chosen to avoid reading meaning into small differences.

Share of policies and share of premium

"Share of policies" is a company's policies in force divided by all policies in TDI's Homeowners line for that year. "Share of premium" is TDI's own market share figure, based on premium dollars for homeowners insurance. The two differ because a company that insures expensive houses collects more premium per policy than one that mostly insures renters.

Rate filings

Insurers must file home insurance rates with TDI before using them. We list homeowners and dwelling filings received in the last five years, matched to companies by exact legal name. The percentage is the insurer's own estimate of the overall effect on existing policyholders. A blank or 0% means the filing changed rules, or something else that does not move premiums overall. One customer's change can be far from the overall figure, and several filings in a year do not simply add up. A filing marked "Reviewed" means TDI finished its review; it does not mean TDI endorsed the rate.

Non-renewals, cancellations and declined applications

A 2025 Texas law, House Bill 2067, took effect on January 1, 2026. It requires insurers to report to TDI every quarter the reasons they declined, cancelled or did not renew policies, by ZIP code, and requires TDI to publish a summary that does not identify any insurer. We use two of the resulting datasets: one with counts by month and ZIP code, and one with the reasons given.

  • Which policies. Unless a page says otherwise, we count owner-occupied homeowners policies only, leaving out renters, condo, rental dwelling and mobile home policies.
  • Cancellations are mostly not the insurer's decision. The largest reasons reported are "at the insured's request" and non-payment. We lead with non-renewals because they are the insurer's choice.
  • Declined applications count applications, not homes. One household shopping with five insurers can be declined several times.
  • Reasons can overlap. A notice can list more than one reason, so shares can total more than 100%.
  • The record is short. Only a few months have been published so far. That is enough to compare places, not to show a trend.

Non-renewals per 1,000 policies

This rate is our own arithmetic: non-renewals of owner-occupied homeowners policies over the reported months, divided by homeowners policies in force in the same area, times 1,000. The two numbers come from separate TDI collections that may not define a homeowners policy identically, so treat the rate as an approximation for comparing one place with another. We do not show a rate for a ZIP code with fewer than 200 homeowners policies, and the statewide ranking includes only ZIP codes with at least 2,000.

ZIP codes, cities and counties

TDI reports by ZIP code. ZIP codes are mail delivery routes and do not follow city limits or county lines. Using Census Bureau files, we list a ZIP code under a city when at least a quarter of the ZIP code's land lies inside the city limits, or when the ZIP code contains at least a tenth of the city's land. Some homes in a ZIP code will therefore be outside the city, and a ZIP code can appear under two neighboring cities. For county totals of non-renewals, each ZIP code is assigned to the one county that holds most of its land. County policy counts, by contrast, are TDI's own.

First tier coastal counties

The list of 14 first tier coastal counties comes from the Texas Insurance Code, section 2210.003. A small part of Harris County is also in the Texas Windstorm Insurance Association's coverage area; we do not flag Harris County cities as coastal because most of the county is outside it.

What the site does not have

  • Premiums for a city, a ZIP code or an insurer. TDI publishes statewide averages and a county map. We do not estimate anything finer than that. For sample rates by company, use HelpInsure.com.
  • Financial strength ratings. They belong to the rating agencies and change over time.
  • Which insurer is non-renewing where. The law keeps insurers anonymous in that data.
  • Claims outcomes. Complaints are not claims, and TDI does not publish claim denial rates by company.

Questions and corrections

Write to hello@pickmyhomeinsurance.com. If you think a TDI number is wrong, TDI is the source to contact; if you think we copied or calculated something wrongly, we want to know.