– B –
Glossary and Abbreviations
2010 Demonstration Data Products (DDP): A series of data files released by the Census Bureau, beginning in late 2019, that replicate some of the content of previously published 2010 Census data tables but that are generated by applying the 2020 Census Disclosure Avoidance System to the original 2010 Census data.
ACDEB: Advisory Committee on Data for Evidence Building
ACO: Area Census Office
ACS: American Community Survey
Active Block Resolution: In the original plan for In-Office Address Canvassing for the 2020 Census, a program that permitted Census Bureau analysts to virtually canvass “active blocks” flagged by Interactive Review through available information resources, such as state and local government property databases and commercial street-view imagery. In-office analysts could document and correct housing unit coverage problems based on the available evidence, permitting those blocks to be reclassified as “passive” and kept out of the In-Field Address Canvassing workload. Active Block Resolution began at the Census Bureau’s National Processing Center in spring 2016 but was discontinued in early 2017 due to budgetary constraints.
AdCan: Address Canvassing
AdRec; AdRecs: administrative records
Address Canvassing (AdCan): A major census operation to improve the completeness of the Master Address File, conducted in the year before the 2000, 2010, and 2020 Censuses. The operation is particularly focused on
addresses that will be included in the main census mailings, that is, census blocks in the Mailout/Mailback (2000 and 2010) and Self-Response (2020) type of enumeration area. The 2000 Census Block Canvassing operation and the 2010 Census Address Canvassing operation were both intended to send field staff to every address and street in every block in the designated areas, with the 2010 Address Canvassing operation being the only 2010 operation to utilize handheld computers for data collection (thus acquiring map-spot locations for addresses whenever possible). The 2020 Census Address Canvassing operation had both In-Office and In-Field components, with the In-Office program further divided into several subcomponents (most prominently, Interactive Review and Active Block Resolution) until some of the subcomponents were discontinued for budgetary reasons. In support of the 2020 Census, In-Field Address Canvassing operated between August–October 2019, with temporary-hire listers carrying out the field operation using laptop computers (and not the smartphones used in later 2020 Census field operations).
Administrative records (AdRec; AdRecs): Records collected as part of the operation of federal, state, and local programs, typically fund allocation and tax programs. A prominent part of the research program of the 2010 and 2000 Censuses, administrative records were given wider use in the 2020 Census, both for the identification of probable vacant housing units and for the enumeration of occupied housing units (in lieu of multiple field visits in some cases and as a last resort in others).
Age heaping: Phenomenon in censuses and surveys in which respondent age is reported at certain levels at the expense of other levels; in U.S. experience, age heaping typically refers to excess reporting of round-sounding ages ending 0 or 5.
AIAN: American Indian or Alaska Native
American Community Survey (ACS): The U.S. Census Bureau’s continuous survey program to measure detailed socioeconomic and other data that now operates in parallel with the decennial census. In the 2010 Census, the ACS formally replaced the long-form sample of households that had been administered a more detailed questionnaire in previous decades, making the 2010 and 2020 Censuses short-form only.
Apportionment: The reallocation of the 435 seats in the U.S. House of Representatives, based on population size, using data from a new decennial census. Under current law, apportionment is conducted using the method of equal proportions.
Area Census Office (ACO): In the 2020 Census, a temporary office established to support decennial census field operations, in turn overseen by one of the temporary Regional Census Centers; in the 2010 Census, these offices were known as Local Census Offices. The 2020 Census used a field infrastructure consisting of 248 Area Census Offices divided between 6 Regional Census
Centers, generally co-located in cities containing permanent Census Bureau Regional Offices (Atlanta, Chicago, Los Angeles, New York, and Philadelphia, with the Denver region’s Regional Census Center being established in Dallas). By comparison, the 2010 Census used 494 Local Census Offices across 12 Regional Census Centers (the Census Bureau consolidated to six operational regions in 2012).
ASA: American Statistical Association
Basic Collection Unit (BCU): The smallest unit of geography used in 2020 Census data collection. Previous censuses designated separate “collection blocks” and “tabulation blocks,” the former being used internally during data collection and the latter being the final, published identifiers for census blocks. The BCU concept largely inherited its form and labels from the vintage of census blocks (or block clusters) then in use by the Census Bureau for census testing and other survey programs. Importantly, type of enumeration area is assigned to each BCU.
BCU: basic collection unit
CBAMS: Census Barriers, Attitudes, and Motivators Survey; family of Census Bureau experiments in recent decennial censuses to elicit information on factors driving response and nonresponse to the census.
CCF: Coverage Correction Factor
CEDCaP: Census Enterprise Data Collection and Processing
CEF: Census Edited File
Census Edited File (CEF): The third major step in the product flow for post data collection processing in the 2020 Census, following the Census Unedited File. The CEF includes additional edits and imputation (affecting the characteristics of persons or housing units but not the overall counts), and is effectively the complete data set for all items in the census. The CEF is processed by the Disclosure Avoidance System to protect confidentiality of census responses; in the 2020 Census, it is the resulting Microdata Detail File that is used for tabulation of census products, rather than the CEF.
Census Field Supervisor Area (CFSA): Operational level of geography developed to distribute field staff work. In the 2020 Census Nonresponse Followup operation, changes in phase between full automation, semipermanent assignment, and closeout (phases 1–3 in the new structure for the program) were determined by case completion rates at the CFSA level. As a purely operational, not cartographic, geography, CFSAs do not correspond neatly with other similar geographies such as tracts.
Census Program of Evaluations and Experiments (CPEX): In recent U.S. censuses, the slate of formal operational assessments, evaluation reports, and major experiments conducted in conjunction with the census. The set of census coverage measurement activities (Demographic Analysis and the Post-Enumeration Survey)—and their results—are not themselves generally
under the CPEX umbrella, but the operational conduct of specific Post-Enumeration Survey suboperations is subject to CPEX study.
Census Review Analysis and Visualization Application (CRAVA): Data visualization and analytical software developed by the Census Bureau to permit its staff and subject matter experts to review and correct problems during the post data collection processing stage of the census (i.e., progressing from Decennial Response File to Census Unedited File to Census Edited File). It is no longer in use.
Census Questionnaire Assistance (CQA): In the 2020 Census, the operation responsible for telephone-based operations in the census. As the name implies, a main focus of the operation was a dial-in line through which census respondents could obtain information about completing the census form and request a paper questionnaire or a foreign language questionnaire/language assistance guide. Callers to the CQA line were asked if they wanted to complete the census questionnaire during the call; if so, CQA agents conducted the census interview (thus making CQA the source for telephone-based Self-Response to the census). In response to the COVID-19 pandemic, CQA functionality was expanded to include some provision for outbound CQA calling (e.g., to try to complete Nonresponse Followup cases where physical access was difficult).
Census Unedited File (CUF): The second major step in the product flow for post data collection processing in the 2020 Census, following the Decennial Response File 2. The CUF reflects the application of household count imputation and so represents the final universe of address and population counts; accordingly, the state-level apportionment counts are generated from the CUF.
CFSA: Census Field Supervisor Area
CNSTAT: Committee on National Statistics of the National Academies of Sciences, Engineering, and Medicine
Count Question Resolution (CQR): An operation in the 2020 Census and other recent censuses in which state, local, and tribal governments can challenge their population totals if the officials can demonstrate flaws in the geographic boundaries used to generate population counts or in the location (or associated population count) of housing units. It is a postcensus review activity focused only on housing units, and so differs in scope from the new-in-2020 Post Census Group Quarters Review program; it also differs from the Count Review program in the late stages of conducting the census, which engages state-level participants to review housing unit counts for general accuracy.
Count Review (CR): Late-stage census operation in which state-level members of the Federal-State Cooperative for Population Estimation are permitted to review preliminary counts near the end of data collection, to suggest necessary updates or revisions to the Master Address File.
Coverage Correction Factor (CCF): See Coverage Measurement.
Coverage Measurement:1 A primary means of measuring census quality by estimating undercount and overcount for the nation as a whole and for specific demographic groups. The set of census enumerations from a sample of clusters of census blocks (known as the E-sample) is matched to the results of the independent Post-Enumeration Survey (PES) conducted in the same sample of block clusters, and the statistical technique of dual-systems estimation is used to estimate under- or overcount. In brief:
- Every enumeration recorded in the census (and hence the E-sample) is either “data-defined” (containing sufficient information to enter census processing and editing) or a “whole-person imputation” (in which every person characteristic had to be imputed).
- Enumerations that are not whole-person imputations may be divided into two groups, “correct enumerations” and “erroneous enumerations,” with the latter principally consisting of duplicate entries.
- In this way, the major components of census coverage are correct enumerations, erroneous enumerations, whole-person imputations, and omissions.
- Initial Coverage Correction Factors (CCFs) for a particular population group are obtained by multiplying the estimated probability that a census enumeration is data defined by the estimated probability that the enumeration is correct, and dividing by the estimated probability that the enumeration is a match between the E- and P-samples. (Some CCFs are further adjusted based on Demographic Analysis sex ratios to account for known, systematic effects.) The sum of the CCFs for the population group is a synthetic estimate of the size, which is then compared to the census count to estimate net coverage rate
CPEX: Census Program of Evaluations and Experiments
CQA: Census Questionnaire Assistance
CQR: Count Question Resolution
CR: Count Review
CRAVA: Census Review Analysis and Visualization Application
CSAC: Census Scientific Advisory Committee
CUF: Census Unedited File
DA: Demographic Analysis
DAS: Disclosure Avoidance System
___________________
1 In the 2020 Census, Coverage Measurement is the heading used to describe component operations related to this work, though the underlying survey (and hence the program as a whole) is commonly termed the Post-Enumeration Survey (PES). In the 2010 Census, the corresponding suite of operations was known as Census Coverage Measurement (CCM); in the 2000 Census, it was the Accuracy and Coverage Evaluation (A.C.E.).
DDHC; DDHC-A; DDHC-B: Detailed Demographic and Housing Characteristics Files
DDP: 2010 Demonstration Data Products
Decennial Response File (DRF): The first major step in the product flow for post data collection processing in the 2020 Census, in two stages. The Decennial Response File 1 (DRF1) is the general compilation of all responses by all modes, including duplicate responses; Decennial Response File 2 (DRF2) reflects unduplication of census returns across households using the Census Bureau’s proprietary Primary Selection Algorithm.
Delivery Sequence File (DSF): The master database of deliverable mail addresses maintained by the U.S. Postal Service, organized by carrier route. First used as a source of address updates to the Census Bureau’s Master Address File for the 2000 Census, regular “refreshes” from the DSF remained a primary source of new address information in the 2010 and 2020 Censuses.
Demographic Analysis (DA): A census coverage measurement technique that derives population estimates of national and major demographic subgroups (that are independent of the current census) based on administrative records data on births, deaths, immigration, and emigration. Demographic Analysis estimates have been extended back to the 1940 Census.
Demographic and Housing Characteristics Files (DHC): Series of 2020 Census data products largely replacing the content of previous decades’ Summary File releases.
DHC: Demographic and Housing Characteristics Files
Differential privacy (DP): A system for sharing information that imposes quantifiable limits on the disclosure of information about individual records in the database, as through the injection of statistical noise into all data cell entries. In the 2020 Census, the Census Bureau decided to implement a new Disclosure Avoidance System (DAS) premised on differential privacy theory; the TopDown Algorithm is the specific differentially private (or formally private) algorithm used to generate the first major 2020 Census data products after the state-level apportionment populations: namely, the redistricting data required under P.L. 94-171 and the forthcoming Demographic and Housing Characteristics Files. See also privacy-loss budget.
Disclosure Avoidance System (DAS): The intermediary system between the Census Edited File and the final data files used for tabulation of census products that applies techniques to curb the risk of identification of census respondents. For the 2020 Census redistricting data, the 2020 Census DAS uses the TopDown Algorithm to take differentially private “noisy measurements”—crosstabulation cells with a controlled amount of statistical noise added to them—and convert them into a synthetic Microdata Detail File that is used for tabulation purposes.
DOB: date of birth
DP: differential privacy
DRF; DRF1; DRF2: Decennial Response File
DSE: dual-systems estimation
DSF: Delivery Sequence File
Dual-systems estimation (DSE): Statistical procedure for estimating the size of a population by combining information from two independent listings/samples of that population. See coverage measurement.
E-sample: In dual-systems estimation, the set of decennial census (enumeration) returns for the same set of blocks in which the independent Post-Enumeration Survey is conducted.
ECaSE-ISR: Enterprise Census and Surveys Enabling–Internet Self-Response
Enumeration at Transitory Locations (ETL): The operation in the 2020 Census and other recent censuses that counts the population at sites where the location itself or residence in individual units is inherently transitory or impermanent (that is, where people are unlikely to live year-round). A key characteristic of transitory locations is that people either pay a fee to stay there (hence the 2020 Census transitory location categories of campgrounds, recreational vehicle (RV) parks, marinas, and hotels/motels) or work there on an impermanent basis (hence, the categories of racetracks and circuses/carnivals). Transitory locations are sometimes confused with group quarters, and there is also potential overlap between transitory units (individual units or spaces in transitory locations) and housing units, with an occupied transitory unit being considered a housing unit if at least one person usually lives there. The ETL operation is preceded by a Transitory Location Advance Check operation to broker arrangements for the count.
EPIK: enhanced protected identification key
∊ (epsilon): See privacy-loss budget.
eResponse: Response mode in the 2020 Census Group Quarters Enumeration operation through which group quarters facility administrators could enter response data for all residents as of Census Day in a Microsoft Excel template provided by the Census Bureau and submit it through a secure internet portal.
Erroneous enumeration: See coverage measurement.
ETL: Enumeration at Transitory Locations
FCSM: Federal Committee on Statistical Methodology
FERPA: Family Education Rights and Privacy Act
Field Verification (FV): Field operation that occurs as part of Nonresponse Followup and uses the same enumerators, but in which the focus is solely on verification of housing unit addresses newly detected during census data collection and collection of geographic coordinates for the address (not the conduct of an interview with household respondents). In the 2020 Census, FV cases were those addresses originating from Non-ID Processing that
could not be automatically matched to the Master Address File or resolved in In-Office Address Verification.
FSCPE: Federal-State Cooperative for Population Estimates
FV: Field Verification
FY: fiscal year
GAO: U.S. Government Accountability Office
GQ: group quarters
GQAC: Group Quarters Advance Visit
GQE: Group Quarters Enumeration
Group quarters (GQ): A place where people live that is not a housing unit. Group quarters are commonly categorized as institutional (e.g., correctional facilities and nursing or other health care facilities) or noninstitutional (e.g., college student housing, military barracks, ships, group homes, worker dormitories, and shelters). GQs differ from housing units because GQs are owned or managed by some entity providing housing or custodial/medical services to the residents; they differ from transitory locations because people commonly pay a fee for an impermanent stay at a transitory location.
Group Quarters Advance Visit (GQAC): The census field operation through which initial contact is made with a group quarters site, basic information is collected to properly categorize the site’s group quarters type, and arrangements are made for data collection in the Group Quarters Enumeration operation.
Group Quarters Enumeration (GQE): The census field operation through which data is collected from or about group quarters residents, whether through administration of individual questionnaires or provision of data from the facility as a whole (e.g., in the 2020 Census, through the eResponse internet response option for group quarters). Emergency and transitional shelters, soup kitchen and regularly scheduled mobile food van locations, and targeted non-sheltered outdoor locations are all considered group quarters in the decennial census framework, and so the Service-Based Enumeration operation that covers those locations that may be used by people experiencing homelessness is an operational adjunct to the broader GQE.
GSS-I: Geographic Support Systems Initiative
HUD: U.S. Department of Housing and Urban Development
IFAC: In-Field Address Canvassing
In-Field Address Canvassing (IFAC): See Address Canvassing.
In-Office Address Canvassing (IOAC): See Address Canvassing; Active Block Resolution; Interactive Review.
INR: item nonresponse
iCADE: The Integrated Computer Assisted Data Entry system to scan paper questionnaires and capture the data in electronic format for further
processing. The iCADE system was developed within the Census Bureau and has been used in various Census Bureau surveys and some international censuses. Paper data capture was outsourced in the paper-based 2010 Census, but the in-house iCADE system was adopted to process paper questionnaires in the 2020 Census.
Interactive Review: Lead portion of the 2020 Census In-Office Address Canvassing program, beginning in fall 2015 and continuing through early 2019, in which analysts viewed Master Address File housing unit locations and counts superimposed on aerial imagery to identify change or stability in residential structures. Interactive Review was meant as a comprehensive but quick triage, classifying blocks as active (evidence that the number of housing units visible in imagery differs from the number in the Master Address File), passive (stable blocks, in terms of housing unit numbers in both sources), and on hold (blocks for which the available imagery was inconclusive). In the original program, active blocks were sent to the Active Block Resolution stage; after that stage was discontinued, blocks unresolved in Interactive Review were sent to In-Field Address Canvassing (though a number of “trigger” events, such as the availability of updated imagery, could prompt another round of Interactive Review for blocks).
Internet Choice; Internet First: The two principal mail contact strategies assigned to census tracts in the Self-Response type of enumeration area based on estimates of likelihood to respond via the internet, as derived from American Community Survey (ACS) data. Internet Choice areas, with older populations and low internet take-up in the ACS, received a paper questionnaire in the initial mailing (along with instructions for responding online); Internet First areas, deemed more likely to readily respond via the internet, received multiple letter/postcard cues to respond online before a paper questionnaire was sent to still-nonresponding households.
Internet Self-Response (ISR): A designated operation of the 2020 Census to facilitate online response to the census. Core to the operation was the development of an internet response platform—first the Census Bureau-developed Primus system, then the outsourced eCaSE-ISR system, and ultimately a return to Primus.
IOAC: In-Office Address Canvassing
IRS: Internal Revenue Service
IT: information technology
JASON: An independent group of scientists engaged to advise the federal government on science and technology matters, principally in the defense and intelligence arenas. JASON administratively operates through the MITRE Corporation; despite the capitalization, JASON is not an acronym.
Local Update of Census Addresses (LUCA): A Census Bureau program developed in response to the Address List Improvement Act of 1994 (and so included in the 2000–2020 Censuses) in which local officials were given
the opportunity to review individual addresses on the Master Address File and make corrections, additions, and deletions to that list, and to make corrections to census maps to match any changes that may be needed.
LUCA: Local Update of Census Addresses
MAF: Master Address File
Master Address File (MAF): The master list of residential and nonresidential addresses maintained by the Census Bureau. Since the 2000 Census, the MAF has been maintained as an ongoing resource for the Census Bureau’s censuses and surveys, updated using a variety of sources including the U.S. Postal Service’s Delivery Sequence File and, in the decade preceding the 2020 Census, a major Geographic Support System initiative to obtain address and geographic files from state, local, and tribal governments. MAF entries are linked to a specific geographic location (for eventual tabulation purposes) using the Census Bureau’s principal cartographic resource, the Topologically Integrated Geographic Encoding and Referencing (TIGER) System. Extracts are derived from the MAF using a series of filters to support specific operational purposes; the report text examines two of these extracts, denoted MAF1 and MAF2, that correspond to the set of addresses deemed workable for enumeration at the beginning of 2020 Census operations and the set of addresses enumerated in the 2020 Census, respectively.
MAFID: Master Address File ID (identifier)
MAPE: mean absolute percentage error
MENA: Middle Eastern or North African
Metadata: Data about other data, including quality metrics data (e.g., item nonresponse rates), data on the progress of individual operations in a census or survey, or paradata concerning the manner of data collection.
Middle Eastern or North African (MENA): An additional major race category that has periodically been suggested for addition to the federal standards in Statistical Policy Directive No. 15.
Mobile Questionnaire Assistance (MQA): An adjunct to the 2020 Census Internet Self-Response operation, added very late in the planning process in response to Congressional appropriations language. MQA was intended to deploy temporary Questionnaire Assistance Centers at selected, large public gatherings in areas with low self-response rates. Program staff with Census Bureau-issued mobile devices would allow visitors to complete their census form online. However, the operation was largely scaled back due to COVID-19-related prohibitions on public gatherings.
Modified Race File: A special tabulation of census data that allocates Some Other Race Alone responses (i.e., in which the only major race category selected was Some Other Race) to other specific race groups, based on the ethnicity question and other factors. This file is necessary for detailed Demographic Analysis results by race because vital records and other data
sources do not include Some Other Race as an option; production of the Modified Race File has been greatly slowed in the 2020 Census for lack of a disclosure avoidance routine consistent with the differential privacy-based 2020 Disclosure Avoidance System.
MQA: Mobile Questionnaire Assistance
MWI: Modified Whipple’s Index
National Processing Center (NPC): The Census Bureau’s principal facility for collecting and processing the returns from its censuses and surveys, as well as for providing logistical and clerical support for decennial census field operations. The NPC is located in Jeffersonville, Indiana. In the 2010 Census, the NPC housed a data capture center that shared paper data capture responsibilities with two other centers (in Essex, Maryland, and Phoenix, Arizona) that were administered by external contractors; in the 2020 Census, paper data capture was shared between the NPC and a facility in Phoenix.
NC: New Construction
New Construction (NC): A census geographic partnership and address-building operation in which state, local, and tribal governments submit new mailing addresses for living quarters under construction but that were expected to be complete by Census Day.
Non-ID Processing: The system in the 2020 Census to facilitate the collection of respondent information (including self-supplied usual residence information) without requiring the respondent to enter a Master Address File-keyed identifier code from a physical mailing to a household. In the 2010 Census, Non-ID nonreturns were primarily the blank-address “Be Counted” forms made available in public places for people to complete if they believed they had been missed in the census. However, in the 2020 Census, Non-ID Processing was critical to the extensive promotion of the notion of replying to the census anytime or anywhere, using the Internet Self-Response option. The address information collected in Non-ID Processing is compared and reconciled with the Master Address File, with ungeocoded addresses being designated for verification by field staff.
Nonresponse Followup (NRFU): The census field operation whereby census enumerators attempt to collect responses from households for which no response was obtained by Self-Response methods (in the 2020 Census, by Internet, mail, or phone) or returning a questionnaire left with the housing unit in Update Leave. In the 2010 Census, NRFU was performed exclusively using paper questionnaires, with assignments parceled out to individual enumerators through a hierarchy of field staff. The 2020 Census utilized electronic interviewing on smartphones, with assignments pushed out to enumerators’ devices on a nightly basis; particularly in the early stages, the enumerator was also given an optimized route to follow in conducting the interview attempts. NRFU in the 2020 Census was also
the first in which the availability of reliable administrative records data for a household could be used for enumeration in lieu of second and subsequent enumerator visits.
NPC: National Processing Center
NRFU: Nonresponse Followup
NRFU RI: Nonresponse Followup Reinterview; a quality assurance operation
Numident: Short for Numerical Identification System, the database maintained by the Social Security Administration for assignment of Social Security numbers and registration in the system.
OBAV: Office-Based Address Verification
Office-Based Address Verification (OBAV): An operation in the 2020 Census in which Census Bureau office staff worked to resolve addresses coming in through Non-ID returns (i.e., without reference to the printed identifier number in any census mailing or delivery) that did not readily match to the Master Address File, without requiring a field visit.
OMB: U.S. Office of Management and Budget
P-sample: Set of returns from the Post-Enumeration Survey, a sample survey of respondents in a sample of block clusters conducted after (and independent of) the census enumeration; part of the Census Coverage Measurement program, it is used to calculate the match rate in the dual-systems estimation formula.
Paradata: Operational or functional data from a census or survey that describe the manner in which the census/survey data were collected, such as the number of contacts made with a survey respondent or the mode of data collection. Paradata is essentially a subset of metadata.
PCGQR: Post Census Group Quarters Review
Person Identification Validation System (PVS): The Census Bureau computer system for assigning protected identification keys to administrative records and other data sources to facilitate data linkage.
PES: Post-Enumeration Survey
PIK: protected identification key
Post Census Group Quarters Review (PCGQR): A special review program permitting state and local governments to challenge and correct the 2020 Census group quarters enumeration in their jurisdictions; though the results do not change the official 2020 Census totals (for redistricting and other purposes), it allows for correction of annual population estimates going forward.
Post-Enumeration Survey (PES): Both a generic term for the independent, follow-up survey conducted in census coverage measurement programs based on dual-systems estimation and the specific operational name given to said survey in some decennial censuses (as done in 2020).
Primus: The technical platform for Internet Self-Response in the 2020 Census. Primus was developed in-house by the Census Bureau in early 2020 Census
planning and testing but was shifted to backup-system status in mid-decade in favor of an externally developed software solution. Immediately prior to the 2020 Census, the Census Bureau concluded that Primus was better able to handle high volumes of concurrent internet users, returning it to use as the primary response channel in the census.
Privacy-loss budget: The parameter in a differential privacy implementation, technically defined as the maximum distance between queries performed on two databases that differ from each other by one entry, that is conceptualized as the amount of information disclosure risk that will be tolerated. The privacy-loss budget is commonly denoted by ∊ and characterized as a slider between privacy and accuracy, ∊ → 0 meaning perfect privacy and zero accuracy and ∊ → ∞ connoting absolute accuracy and zero privacy. The Disclosure Avoidance System allocates certain shares of overall ∊ to specific histogram cells and levels of geographic tabulation, thus determining the amount of statistical noise added to the cells.
Protected identification key (PIK): The anonymized identifier variable assigned to entries in administrative records source files and other data resources to enable record matching across sources.
Proxy: A census interview in which the respondent is not a member of the household being enumerated. The respondent might be a neighbor, a landlord, or some other knowledgeable person.
PVS: Person Identification Validation System
Regional Census Center (RCC): See Area Census Office.
ROCkIT: Re-Organize Census with Integrated Technology
SafeTab-H; SafeTab-P; PHSafe: Differential privacy-based algorithms developed for generating the Detailed Demographic and Housing Characteristics Files and other late-stage 2020 Census data products.
SAIPE: Small-Area Income and Poverty Estimates
SBE: Service-Based Enumeration
Service-Based Enumeration (SBE): In the 2010 and 2020 Censuses, the component of the broader Group Quarters Enumeration program focused on people experiencing homelessness who could be reached at service or shelter locations or in certain targeted non-sheltered outdoor locations. In 2010, the program was spread across three days around the April 1 reference date, each day specific to a type of location; in the 2020 Census, SBE was postponed until September 22–24.
SF1; SF2: Summary File 1; Summary File 2
Some Other Race (SOR): The sixth major race category permitted by the Census Bureau for response to the race question. It is not defined or required by the U.S. Office of Management and Budget’s Statistical Policy Directive No. 15 but is rather required by Congress through language in Census Bureau appropriations.
SOR: Some Other Race
SRQA: Self-Response Quality Assurance
SSN: Social Security number
Statistical Policy Directive No. 15: Common name for the 1977 version of standards issued by the U.S. Office of Management and Budget for the collection of data on race and ethnicity (Hispanic origin). The most recent revision (1997) of these guidelines is not labeled or numbered as a directive or circular. The guidelines make it possible to identify with more than one race, rather than choosing from only one of a few mutually exclusive race categories.
Summary File 1 (SF1); Summary File 2 (SF2): The detailed data files of tabulations that were the primary data products of recent U.S. censuses, until replaced in the 2020 Census by the Demographic and Housing Characteristics Files.
Targeted non-sheltered outdoor locations (TNSOL): Technically designated as noninstitutional group quarters locations, areas where people experiencing homelessness are known to live without paying to stay, and where they are counted in the Service-Based Enumeration program.
TDA: TopDown Algorithm
TEA: type of enumeration area
Tenure: The status of an occupied housing unit as either owner occupied or renter occupied.
TIGER: Topologically Integrated Geographic Encoding and Referencing (TIGER) System; the Census Bureau’s framework for identifying the exact geographic location of residential addresses (as well as other physical features).
TL: transitory location; see Enumeration at Transitory Locations
TNSOL: targeted non-sheltered outdoor location
TopDown Algorithm (TDA): The differential privacy-based algorithm developed by the Census Bureau to produce the P.L. 94-171 Redistricting Data File and Demographic and Housing Characteristics Files. Generally speaking, the algorithm works separately with person and housing unit data from the Census Edited File, injecting a specified amount of statistical noise into every entry of summary tables/queries (based on allocation of privacy-loss budget to each query) and using nonlinear optimization and other tools on the resulting “noisy measurements” to find a microdata representation that satisfies both noise-injected table values and additional constraints (such as nonnegative counts or totals that must remain invariant to change). The name arises from the manner in which the algorithm generates national-level estimates first, then iterates to produce estimates at the state level, county level, and so forth, down to the census block level.
Transitory location (TL): See Enumeration at Transitory Locations
Type of enumeration area (TEA): A classification developed by the Census Bureau prior to a decennial census to determine the basic strategy used to
conduct the census in particular blocks (or, in 2020 Census terminology, basic collection units). The 2000 and 2010 Censuses each designated nine TEAs, while the 2020 Census utilized six TEAs: Self-Response (subsuming the Mailout/Mailback TEA of the previous censuses), Update Enumerate, Update Leave, Island Areas (separate designation for the U.S. territories), Remote Alaska, and Military.
UAA: Undeliverable as Addressed
UE; UE/RA: Update Enumerate; Update Enumerate/Remote Alaska
UL: Update Leave
Undeliverable as Addressed (UAA): U.S. Postal Service terminology for any mailing that cannot be delivered to the specific name and address on the mailing, for any reason.
Update Enumerate: A method of census enumeration and type of enumeration area in which a census enumerator’s sole visit involves both Address Canvassing functions (adding or updating address list entries) and enumeration (conducting an interview with household respondents). Update Enumerate areas tend to be remote, rural geographic areas; Remote Alaska areas are counted using Update Enumerate methods but are designated as a separate type of enumeration area because they are approached earlier in the census year (beginning in January) due to physical accessibility concerns.
Update Leave: A method of census enumeration and type of enumeration area, primarily focused on geographic areas where the U.S. Postal Service does not do home delivery or in which there are other possible issues with mail delivery, in which a census enumerator conducts address list update activities in the area and leaves a questionnaire package at each housing unit. Update Leave was not originally planned to be a type of enumeration area in the 2020 Census, as it was in 2010, but it was reinstated in the 2020 Census plan in 2017. In addition to areas lacking mail delivery, Update Leave has also been applied to areas that have been affected by major disasters prior to the census.
USPS: U.S. Postal Service
Whole-person imputation: A census enumeration for which all characteristics information is unknown and requires imputation, occurring when a household is believed to be occupied but no response data of any kind is available.
zCDP: Zero-concentrated differential privacy; a variant of pure differential privacy that relaxes some assumptions. Beginning April 2021, the Census Bureau’s TopDown Algorithm was reconfigured to use discrete Gaussian noise infusion and zCDP in lieu of discrete Laplace noise infusion and pure differential privacy.
