SOI Tax Stats - Estate Tax Study Data Sources and Limitations


Return to the main Estate Tax Study Metadata page

This page contains information about data sources and limitations for SOI's Estate tax study.

Please visit Estate Tax Statistics to access articles and tables from the study.

Analysts in the Estate Tax Section of SOI's Individual and Tax-Exempt Branch, collaborating with SOI staff in the Kansas City Submission Processing Center, conduct the annual Estate Tax Study, which extracts demographic, financial, and bequest data from United States Estate (and Generation-Skipping Transfer) Tax Return (Form 706), the Federal estate tax return. The study produces filing year data on estate taxation, focuses on a single year of death for a period of 3 years, and also provides periodic year-of-death estimates.

Year-of-death estimates are advantageous in that the included estates are therefore subject to the same tax law and similar economic conditions. A single year of death is sampled for 3 calendar years, and 99 percent of all returns for decedents who die in a given year are filed by the end of the second calendar year following the year of death. The Estate Tax Study for the period 2016–2018 concentrates on Year of Death 2016, the most recent year-of-death estimates available.

For each study year, 2016–2018, a sample was selected from returns filed. The table below provides the number of returns sampled for each year of death, the total population estimates for those years, and the corresponding number of 2016 decedents..

Filing Year Returns Sampled Population Estimates 2016 Decedents
2016 7,959  12,411 1,012
2017 8,060  12,711 9,982
2018 7,724  13,526 2,435 
Total 23,743 38,648 13,429

Estate tax returns were sampled while the returns were being processed for administrative purposes, but before any audit examination. Returns were selected on a flow basis, using a stratified random probability sampling method, whereby the sample rates were preset based on the desired sample size and an estimate of the population. The design had three stratification variables: year of death, age at death, and size of total gross estate plus adjusted taxable gifts. For Filing Years 2016–2018, the year-of-death variable was separated into two categories: 2016 year of death and non-2016 year of death. Age was disaggregated into three categories: under 65, 65 under 80, and 80 and older (including age unknown). Total gross estate plus adjusted taxable gifts was limited to four categories: under $5.0 million, $5.0 million under $10.0 million, $10.0 million under $20.0 million, and $20.0 million or more. Sampling rates ranged from 50 percent to 100 percent; returns for more than half of the strata were selected at the 100-percent rate.

Because almost 99 percent of all returns for decedents who die in a given year are filed by the end of the second calendar year following the year of death, and because the decedent's age at death and the length of time between the decedent's date of death and the filing of an estate tax return are related, it was possible to predict the percentage of unfiled returns within age strata. As a result, the sample weights were adjusted accordingly, in order to account for returns for 2016 decedents not filed by the end of Filing Year 2018.