Document Type


Publication Date



Oregon -- Population -- Statistics, Demographic surveys -- Oregon, Population forecasting -- Oregon


The methodology used to produce population estimates as of 4/1/2020 for the state at the block group level relies on data from the 2010 Census Summary File 1 (SF1), 2020 Census Apportionment Results, US Census partnership programs supporting the 2020 Census, 2020 Tiger/Line boundary files, 2015-2019 5-year American Community Survey (ACS) files, ESRI Updated Demographics vintage 2020, and estimates from the Portland State University Population Research Center and US Census Population Division.

The 2010 decennial census block file was the baseline for average household size, and counts of housing units and population by race in 2010. Block group or tract level trends were measured and then statistically significant differences from the most recent available data were used to project forward household size for constituent blocks when significant differences existed in one block group over time. When the ACS was the basis of comparison, ACS data from the 2006-10 5-year summary file was used, to ensure similar methodology and definitions used in the comparison.

Vintage 2019 Address Count listing data on housing units per block were used to estimate initial new population per block. Updated group quarters population estimates were tallied and subtracted from estimates when total populations only were reported to generate household population estimates, such as for the statewide 2020 apportionment and residential population total. GEOIDs for summary level 157 (incorporated place within county) and 795 (public use microdata area or PUMA) were added to census blocks to enable reweighing to independent estimates. Adjustments were made to ensure consistency between unit counts and population counts when possible (e.g., blocks with one or more housing units in 2019 which had no population in 2010 were imputed average household size and demographics from within the parent block group or tract; blocks with one or more persons in 2010 but no housing units or group quarters in 2019 were set to zero population; etc.).

The 2020 Census Apportionment Results contains total Oregon state resident population of 4,237,256. This includes an estimated 88,649 persons in group quarters living situations. This estimate was used for poststratification weighting to control county estimates to sum to the total state population, according to the population living in households or group quarters. Block group quarters population were controlled to county estimates without intermediate steps. Household populations were controlled top-down from the adjusted county totals to place within county and tracts. Counties with two or more PUMAs had an additional intermediate control step so that the adjusted county total was imposed on the PUMA totals, and then tracts. Extrapolated block population counts from the vintage 2019 address county listing data were controlled to adjusted block group and tract totals, as well as place totals.

The ACS population by race was controlled at the state level to match the state total population, and the ACS population by race by county (from 5-year summary file tables) were controlled to the adjusted county total estimates. The distribution of race for Hispanic/Latino/Spanish origin was adjusted using proportions from the 2010 decennial census, due to differences in methodology and results between the decennial census and the ACS. Below the county level, population by race was controlled top-down in the same way as household populations. Finally, the population by race was converted into proportions of the total population and applied to the previously calculated adjusted block total population (introducing some deviation from ACS race distributions).

To produce estimates in 2020 boundaries, the percentage of built up area in each 2010 block that was within each intersecting 2020 census block was calculated and used to proportionally allocate population from each 2010 block into all related 2020 blocks. This geographic allocation dataset was provided by Amos, Brian, 2021, "2020 Census Block Crosswalk Data",, Harvard Dataverse, V2. Finally, marginal totals were calculated and numbers were rounded to nearest integers. Resampling was performed to ensure that the rounded totals summed to the marginal totals stored prior to rounding.

Note: The Data Dictionary is included below as a supplemental file.


© 2020 Portland State University Population Research Center


Data Description:

Content: - The zip file contains a shapefile (comprised of four total files .shp, .dbf, .shx, and .prj files).

The file can be read in any standard GIS software, or the DBF can be separately read in most statistical software.

Each row consists of one census block, from Tiger/Line vintage 2020 files, in 2020 Census boundaries.

GEOID20 and GEOBK20 fields are standard census block identifiers (new for the 2020 Census). A block GEOID is a concatenation of the state + county FIPS code (5 characters) + census tract code (6 digits) + census block code (4 digits).

The population counts correspond to data elements from forthcoming PL94-171 file tables P1, P2, and P5.

2020 Tabulation blocks can be aggregated to generate other geographies such as districts, cities, etc., by use of the 2020 Block Assignment File (BAF;, and the names corresponding to each geographic code can be found in the Name Look-up Tables (NLTs;

For further information, please contact the Population Research Center: or

Persistent Identifier

Data Dictionary.pdf (38 kB)
Data Dictionary