2020 and 2021 ADI 2020 and 2021 ADI The v4.0.1 ADI has minor refined shrinkage statistical updates to improve the precision of the block group level component estimates derived from the American Community Survey (ACS) data. This results in slight shifts to the block group rankings from the previous version while retaining the general pattern and improving the overall temporal stability of the ADI.
2021 v4.0.1 is based on:
2020 and 2021 ADI The v4 ADI has minor standard shrinkage statistical updates included to mitigate the effect of year-to-year sampling variations in block group level component estimates within American Community Survey (ACS) data. This results in very little actual change in ADI ranking but buffers from known and future expected variation in ACS source data.
2020 ADI The 2020 ADI (v3.2) was constructed using the 2016-2020 5-year estimates from the US Census' American Community Survey. The No Phone variable was removed from the ADI and replaced with a No Household Internet variable to reflect the change in household phone and internet service. The 2020 ADI also is the first version using the redrawn 2020 Census block groups. This results in an increase in block groups from roughly 220000 to roughly 245000. In addition, reapportionment of the block groups results in a number of changes in the boundaries. These geographic boundary changes make comparison impossible between the 2020 and previous versions of the ADI.
2019 ADI The 2019 ADI (v3.1) was constructed using the 2015-2019 5-year estimates from the US Census' American Community Survey. Suppression criteria was updated to use Population, Housing and Group Quarters estimates from the ACS in place of the decennial Census. This adjusts the ADI of approximately 70 block groups nationally.
2019 ADI The 2019 ADI (v3.0) was constructed using the 2015-2019 5-year estimates from the US Census' American Community Survey. The methods include the suppression and imputation criteria initiated with the 2018 version (see entry dated 11/19/2020) and subsequently processed into the 2019 vintage.
2015 ADI The 2015 ADI (v3.0) was constructed using the 2011-2015 5-year estimates from the US Census' American Community Survey. The methods include the suppression and imputation criteria initiated with the 2018 version (see entry dated 11/19/2020) and subsequently processed into the 2015 vintage.
2018 ADI The 2018 ADI (v3.0) was constructed using the 2014-2018 5-year estimates from the US Census' American Community Survey. Two methodological changes have been implemented to address missing Census Block Group level data in the underlying ACS data, survey errors and a new geographically-nested imputation method.
(1) Block group suppression: We have applied the same Diez Roux1 suppression criteria used in the earlier 2013 and 2015 builds: any block group with fewer than 100 persons, fewer than 30 housing units, or greater than 33% of the population living in group quarters will not receive an ADI ranking. In addition, we have suppressed of a small number of block groups which include those with survey errors acknowledged by the US Census Bureau.
(2) Imputation of missing ACS items: We have utilized a geographically-nested imputation methodology consistent with those utilized in other area deprivation indices methodologies, including the English Indices of Multiple Deprivation (IMD) and versions of the Scottish Indices of Multiple Deprivation (SIMD).
The 2015 ADI (v3.0) and 2013 (v2.0) were constructed using the 2011-2015 and 2009-2013 5-year estimates from the US Census' American Community Survey. The methods include the suppression and imputation criteria initiated with the 2018 version and subsequently processed into the 2013 and 2015 vintages.
2015 and 2013 ADI The 2015 ADI (v2.0) was constructed using the 2011-2015 5-year estimates from the US Census' American Community Survey. The methods are identical to the 2013 ADI construction.
Both the 2013 and 2015 ADI now include suppression criteria based on Diez Roux1 which suppress any block group with fewer than 100 persons, fewer than 30 housing units or greater than 33% of the population living in group quarters.