mapping “self-response” for a fair and accurate census · mailing w/paper questionnaire...

18
Mapping “Self-Response” for a Fair and Accurate Census 1 Steven Romalewski Center for Urban Research at the Graduate Center, City University of New York (CUNY) 2020 Census Week 6 Self-Response Analysis (through April 30) Presented on May 1, 2020

Upload: others

Post on 23-Oct-2020

0 views

Category:

Documents


0 download

TRANSCRIPT

  • Mapping “Self-Response” fora Fair and Accurate Census

    1

    Steven RomalewskiCenter for Urban Research at the Graduate Center,

    City University of New York (CUNY)

    2020 Census Week 6 Self-Response Analysis(through April 30)

    Presented on May 1, 2020

  • 2

    • Households that have self-responded as a share (percent) of all known housing units (addresses on the Master Address File) for a geographic area − occupied, vacant, or nonexistent − in the mail-out and Update/Leave universes.

    • Responses can be on-line, by telephone (Census Questionnaire Assistance), or using a paper questionnaire

    • Self-response universe does NOT include: group quarters, transitory locations, military/maritime vessels

    • Rates do not represent percent of people who responded!

    Reminder: what self-response is, and isn’t

  • 3

    OVERALL TRENDS

    • 54.6% nationwide rate as of Thurs., April 30

    • Rate is increasing faster than last week: Two days w/ 0.6 point increase (Apr. 24 & Apr. 29) Overall avg daily increase of 0.3 points this week

    • At this rate, will reach final 2010 rate (66.5%) by June 9 (earlier than we thought per last week)

    • Yesterday was “anniversary” of end of 2010 self-response operation

    (Reminder: this week’s analysis & earlier weeks is available at www.gc.cuny.edu/CUR-research-initiatives )

    Week 6: self-response rates through April 30

    http://www.gc.cuny.edu/CUR-research-initiatives

  • 2020 state-by-state response trends thru April 30

    compared with U.S. rate of increase in 2010*

    *NB: Keep in mind differences in census operations b/w 2010 & 2020 when viewing graph.

  • 5

    States ranked by April 30 response rates

    Top 5Minnesota 65.2 %Iowa 62.5 %Wisconsin 62.1 %Nebraska 61.9 %Michigan 61 %

    Bottom 5Vermont 45.6 %Maine 44.8 %New Mexico 43.1 %West Virginia 42.3 %Alaska 36.3 %

  • 6

    • Internet Choice, Internet First 4th mailing w/paper questionnaire boosting mail

    response Result is most noticeable in Internet First areas Internet Choice rates, on average, still lagging behind

    Internet First

    • Demographic trends Rates in areas with concentrations of historically

    undercounted groups Some areas gained (recent GOTC efforts?), others still

    face challenges

    Key areas of concern analyzed for Week 6

  • 7

    Pace of change: internet response vs mail/phone

    Daily increase frommail/phone exceedsdaily internet increase

  • 8

    Increases in mail response especially noticeable in “Internet First” communities: tract-level examples

    All Internet CHOICEAll Internet FIRST

    Mail/phone rate increased from ~2% to ~13%

  • 9

    Increases in mail response especially noticeable in “Internet First” communities:

    Some county examples for states that are predominantly Internet First

  • Ohio: 76% of housing units received Internet First mailingDaily mail/phone-only response rates shown below for countieswhere share of Internet First exceeds Internet Choice (omitting counties w/higher than avg share of U/L)

    Mail rates start to increase at faster pace (internet rate change stays largely flat)

  • Kansas: 76% of housing units received Internet First mailingDaily mail/phone-only response rates shown below for countieswhere share of Internet First exceeds Internet Choice (omitting counties w/higher than avg share of U/L)

    Mail rates start to increase at faster pace (internet rate change stays largely flat)

  • 12

    • On April 30, the average response rate across all Internet First tracts was 59.1% more than 4 points above U.S. rate (same as 4/23)

    • Average response rate across all Internet Choice tracts was 46.3% 8 points below U.S. rate (last week it was 6 points

    below) response rate gap between Internet First & Choice

    tracts grew to almost 13 points (last week the gap was 10 points)

    (This omits tracts with 10% or more Update/Leave units.)

    Response rates across Internet Choice tracts still lagging behind Internet First

  • 13

    Internet Choice / Internet First response rate gap by city size & bilingual mailing

    Overall Internet First: 59.1%; Internet Choice: 46.3%Gap: 12.8 points

    15.9 pts

    16.5 pts

    18.7 pts

  • 14

    Children under age 5• Tracts with Very High Risk of undercounting young children

    continue to have average response rate below the national level, at 48.8%. (Based on tracts analyzed by Population Reference Bureau.)

    People with incomes below the poverty line“High poverty tracts:” more than 30% of residents have incomes below poverty line.• Average response rate across all high poverty tracts is 41.3%,

    more than 13 points lower than national average. (Gap was 10 points as of April 9)

    • Tracts with less than 30% of the population in poverty had an average response rate of 58%, above the national average.

    Historically undercounted populations

  • 15

    Historically undercounted populations People of color(based on single, non-Hispanic race categories; change from earlier analyses)

    NB: Cities in Hawaii are considered unincorporated places and operate under a City/County government structure, so they are not included as “Cities” in this analysis.

    There are only 38 tracts nationwide in which the plurality population is American Indian/Alaska Native (single race, non-Hispanic). Most of these (34) are outside “incorporated places.” Only 4 tracts are in incorporated places, in the smallest city category.

  • 16

    Historically undercounted populations People of color(change in response rates from April 9 to April 30)

  • 17

    • Tracts where a plurality of people age 25+ have bachelor degrees or higher, average tract response rate is 62.8% (8 points above U.S. rate)

    • Tracts where a plurality have a high school degree or less, average tract response rate is 51.5% (3 points below U.S. rate)

    • In large cities (pop. 1 million+), rates for both education groups are lower: Tracts w/plurality of bachelors degree or higher have avg rate of 52.6%

    (almost at U.S. rate) Tracts w/plurality of HS degree or less have avg rate of 42.5% (10 points

    below higher-degree tracts, and 12 points below U.S. rate)

    Response rates by education level For people age 25 or older

  • Contact for questions, additional information:

    18

    www.CensusHardtoCountMaps2020.us

    Steven RomalewskiCenter for Urban Research at the

    Graduate Center, [email protected]

    http://www.censushardtocountmaps2020.us/mailto:[email protected]

    Mapping “Self-Response” for�a Fair and Accurate CensusSlide Number 2Slide Number 3Slide Number 4Slide Number 5Slide Number 6Slide Number 7Slide Number 8Slide Number 9Slide Number 10Slide Number 11Slide Number 12Slide Number 13Slide Number 14Slide Number 15Slide Number 16Slide Number 17Contact for questions, additional information: