Revised 2019 SPM Reference Year Variables in the 2020 ASEC
Alternate Versions of SPM Variables in 2020 ASEC
Beginning with the 2021 ASEC, the Supplemental Poverty Measure (SPM) variables were estimated with an updated methodology designed to better capture modern living costs and the role of government programs in reducing poverty. These updates affected both SPM resource and threshold variables.
In addition to using the revised methodology in the 2021 ASEC, the Census Bureau released a revised version of the 2019 SPM reference year file (data for the 2019 SPM file was collected as part of the 2020 ASEC1) that uses the updated methodology for estimating SPM measures. IPUMS CPS has integrated this revised version of the 2019 SPM file (collected with the 2020 ASEC) and created alternate versions of variables that differ between the original and revised files. In the CPS ASEC, the SPM reference year is the calendar year to which the income, taxes, transfers, expenses, and poverty thresholds apply.
These new variables, which are denoted with the *_19YRALT suffix, are only available for the 2020 ASEC. The original SPM variables (e.g., SPMWIC, SPMPOV) report the originally released SPM measures for each ASEC when the data were collected. See Table 2 below for a list of alternative SPM variables that differ from the original SPM version
Census and BLS Documentation
- 2019 SPM Research file Readme (2020 ASEC year)
- Improvements to Supplemental Poverty Measure for 2021 (applied to 2019 research file)
- SPM Research Poverty Threshold Overview
Changes to the SPM variables in the 2020 ASEC
| Table 1. Number of Unweighted Persons Classified as In Poverty in the 2020 ASEC | ||
|---|---|---|
| Research File Year | Person Records (unweighted) | |
| 2019 (tabulated using SPMPOV_19RYALT) | 17,848 | |
| 2019 (tabulated using SPMPOV) | 17,697 | |
Resource Adjustments
- The value of the WIC nutrition program was updated to use state-specific average benefit values instead of a single national average, allowing SPM to reflect regional differences in benefits.
Threshold Changes
See this Census Bureau working paper and BLS documentation on Research Poverty Thresholds for more detail.- The inflation adjustment changed from the broad Consumer Price Index for All Urban Consumers (All Items CPI-U) to a composite index "Food, Clothing, Shelter, Utilities, and telephone and internet service" price index; this is referred to as the FCSUti CPI-U.
- The base of thresholds has been moved from the averages within the 30th-36th FCSUti expenditure percentile range to 83 percent of the averages inherent within the 47th-53rd FCSUti percentile range.
- Home internet costs were added as a threshold component.
- Telephone costs were removed from geographic adjustments.
- The estimation sample was expanded from consumer units with exactly two children to all consumer units with children.
- Consumer expenditure data were lagged by one year to estimate the thresholds.
- Certain imputed in-kind benefits (LIHEAP, NSLP, Rental Assistance, and WIC) were added to the thresholds.
Impacts to Data Users
Census Bureau staff did not find statistically significant differences in the overall poverty rate under SPM between the two methodologies. They found that the largest increases under the new methodology were among renters and individuals with disabilities (0.6 and 0.7 percentage points respectively), and that the largest declines in poverty rates occurred for Asians and owners without a mortgage (both 0.4 percentage points). To maximize comparability for analyses using the 2020 ASEC with subsequent years should use the alternative measures for the 2020 ASEC and the primary SPM variables for 2021-onward. Those working with earlier years of data only (e.g., 2011-2020 ASEC samples) should use the primary SPM variables for the most comparable time series.
Footnotes
Supported By
