Weekly Response Rate Analysis

Prepared August 7, 2020

Highlighted topics:

• Selected recent local response rate increases in areas of targeted Get Out The Count (GOTC) efforts;
• Self-response in NRFU “soft launch” areas; what it might mean when NRFU begins nationwide (Aug. 11);
• Areas that have met or surpassed their final 2010 response rates, and areas that are behind their 2010 rates and are most at risk of a rushed NRFU operation;
• Bottom 20 percent of tracts by response rate; and
• Wrapping up; taking stock.

Download the summary presentation of this report

Prepared July 24, 2020

Highlighted topics:

• Areas that have met or surpassed their final 2010 response rates;
• Bottom 20 percent of tracts by response rate:
  • Demographics, compared with top 20%
  • Geographic patterns
  • Implications for Nonresponse Follow-up (NRFU); and
• Update/Leave areas have improved, but rates still low.

Download the summary presentation of this report

Prepared July 24, 2020

Highlighted topics:

• Areas that have met or surpassed their final 2010 response rates;
• Bottom 20 percent of tracts by response rate:
• Demographics, compared with top 20% (geographic patterns, and implications for Nonresponse Follow-up (NRFU); and
• Update/Leave areas have improved, but rates still low.

Download the summary presentation of this report

Prepared June 21, 2020

Highlighted topics:

• Substantial nationwide and statewide (and in Puerto Rico) response rate increases the weekend of June 12;
• Details about those increases in states & communities where Update/Leave operations resumed as of May 6;
• Notable increases also in New York City, Los Angeles, and Chicago;
• Internet First vs Internet Choice: rates increasing for tracts that received bilingual mailing;
• Response rates across tracts by plurality race/Hispanic origin; and
• New online trendline visualization and new response rate metric to help support GOTC work during the extended self-response timeframe.

Download the summary presentation of the report below

---

National Response Rate Trends

The nationwide 2020 Census self-response rate as of Thursday, June 18 was 61.5%. In the past two weeks, the daily increases have been moderate, except for a substantial increase between Friday, June 12, and Monday, June 15, when the U.S. rate increased 0.5 point (before that, the U.S. rate had not increased by more than 0.1 point in one day since May 14 to May 15).

Over the June 12 - 15 weekend, the response rates for several states increased by several points. (Starting June 1, the Census Bureau shifted from reporting daily self-response rates to reporting rates Monday through Friday. Therefore, response rates now reported on a Monday reflect responses from Friday through midnight on Sunday.) Alaska's response rate increased by more than 5 points. Puerto Rico's response rate more than doubled, from 9.2% to 18.9%, and is now 20.3% -- an increase since June 12 of 11.1 points.

In Section 1 of this report, we analyze in more detail the increases in areas covered by the Update/Leave operation. In Puerto Rico, all housing units are included in the Census Bureau's Update/Leave operation, which was suspended in early March due to the pandemic and resumed on May 22. Stateside, most increases between June 12 and June 15 were in states/counties with substantial shares of housing units covered by Update/Leave, while increases since June 15 at the national level have slowed. The notable one-weekend jump in response rates for many Update/Leave areas raised the question of whether there had been a backlog in tracing or checking-in paper questionnaire responses from Update/Leave areas. In fact, the Census Bureau confirmed an initial flaw in the system that tracked mail responses from Update/Leave households, which has now been fixed.

Another notable observation about the overall U.S. response rate is that the share of U.S. response via mail/phone has increased:

• On April 1 (Census Day), 88% of responses were submitted online; 12% via mail/phone.
• By June 18, share of mail/phone responses grew to 20.2% of the U.S. response rate, and online responses decreased to 79.8%.

The response rate trend is shown in the chart in Figure 1, below.

FIGURE 1

tate-by-state response rates

The following charts show the latest 2020 response rates by state and the percentage point change by state from March 20 (when the Census Bureau first reported response rates) through June 18.

Note the inclusion of Puerto Rico in the charts in Figures 2 and 3. Now that the Update/Leave operation has resumed in Puerto Rico, we will be closely tracking self-response rate updates in the Commonwealth. (The entire Commonwealth is covered by the Update/Leave operation.)

Several highlights of statewide response rates include:

• All states except Alaska now have response rates of 50% or more. Alaska has the highest share of housing units covered by the Update/Leave operation (29.8%) of any state.
• Michigan has become the first state to surpass its 2010 response rate of 67.7%; Michigan's response rate on June 18 was 67.8%.
• West Virginia increased its response rate by 4.8 points, from 48.3% on June 12 to 53.1% on June 18. West Virginia has the second highest share of housing units covered by the Update/Leave operation (27.5%) of the 50 states.

FIGURE 2 States ranked by June 18 census response rate (the final 2010 U.S. rate and the current U.S. 2020 rate are included for comparison). 

FIGURE 3 States ordered left-to-right by lowest response rate to highest on June 18 (including the current national rate and final 2010 response rate for comparison). Brown bars represent the beginning response rate (March 20); the blue bars represent the increase between 3/20 and 6/18. 

For Week 13 of our response rate analysis (through June 18), the U.S. response rate increased substantially. But during the prior week of our analysis (Week 12, through June 11), response rate increases in several of the nation's largest cities (with populations of 1 million or more) continued to outpace the nationwide increases. In particular, rates increased at a faster pace than nationwide in New York City, Los Angeles, and Chicago. Rates in these cities also increased notably during Week 13 of our analysis, despite being outpaced by the large U.S. increase the weekend of June 12 to 15; the latter primarily was due to increases in Update/Leave areas, of which few are in cities.

Figure 4 shows the weekly response rate increases (in percentage points) for the U.S. (the black bars) compared with cities with populations of 1 million or more (bars of different color).

FIGURE 4 

Key areas of concern analyzed for Weeks 12 and 13

In addition to these overall response rate trends, we focus our analysis for Weeks 12 and 13 on the following issues:

1. Response rate increases related to Update/Leave operations
2. Internet First vs Internet Choice contact strategies: rates increasing for tracts that received bilingual mailing with paper questionnaire
3. Response rates across tracts by plurality race/Hispanic origin
4. New online trendline visualization and new response rate metric to help support GOTC work during the extended self-response timeframe

1. LATEST RESPONSE RATES RELATED TO UPDATE/LEAVE OPERATION

The Census Bureau's Update/Leave (U/L) operation involves hand-delivering census packets to households in some rural and remote areas, all of Puerto Rico, and on Tribal lands, covering 6.8 million housing units in 2020 (including approximately 1.7 million units in Puerto Rico). Most of these homes (about 95%), however, did not receive their packets before the Bureau suspended the U/L effort soon after it began in mid-March due to COVID-19 restrictions.

On May 6, the Census Bureau began a phased restart of the U/L operation. As of the week of June 8, U/L had resumed everywhere, and on June 18, the Bureau announced that "Census workers have completed 96% of the 2020 Census Update Leave operation."

The map in Figure 5 highlights these states and the Area Census Offices (ACOs) within them that are overseeing the hand-delivery of census packets. The Bureau's website includes a list of all these ACOs, when field operations resumed, and how many housing units in each ACO are covered by the U/L operation.

Note that in some states, the Update/Leave operation did not resume in all Area Census Offices at the same time. Also, the ACOs shown in grey are where hiring is resuming for future census field operations, but these ACOs do not include any housing units covered by U/L. The CUNY HTC/Response Rate map includes a search feature that makes it easy to zoom to any Area Census Office on the map, view the status of the resumption of field operations, and visualize areas within the ACO where the U/L operation has restarted.

FIGURE 5 

In our prior report on response rates in Update/Leave areas (through June 4), we noted that rates in areas with substantial concentrations of housing units covered by the U/L operation were increasing slowly and were being outpaced by increases in non-U/L areas.

These modest increases in U/L areas continued through June 12 (five weeks after the phased restart of U/L had begun). But over the weekend between June 12 and June 15, substantial response rate increases were recorded across the country, especially in areas covered by U/L. The following tables and charts highlight these increases.

Table 1 presents the increases in the top 15 states with the greatest share of estimated housing units covered by U/L (the table includes Puerto Rico, which has 100% of its housing units covered by U/L). The first column shows the increases between June 12 and June 15. The second column presents the increases since June 15 (through June 18), which in several states (such as Wyoming, New Hampshire, and Hawaii) continue to be substantial — as well as Puerto Rico, which increased 1.4 points between June 15 and June 18.

TABLE 1

These increases, however, were not limited only to states with large shares of housing units covered by Update/Leave. Table 2 presents response rates at the tract level, showing rates by state separately for tracts where either most housing units in a tract were covered by U/L, or for tracts with a majority of units covered by the Mail-Out operation. The table only includes states where the response rates in tracts with a majority of housing units covered by U/L increased by 10 points or more between May 6 (when the Census Bureau began to resume the U/L operation) and June 18.

The states in Table 2 are sorted largest to smallest by the increase in response rates for "majority Update/Leave" (U/L) tracts. Notably, the three states with the greatest increase in response rates for majority U/L tracts (New Jersey, Ohio, and Kansas) had some of the smallest concentrations overall of housing units covered by Update/Leave. But the rates in U/L tracts in these states increased substantially.

TABLE

Some visual examples of substantial increases in response rates for small communities covered by Update/Leave are shown below in Figures 6 through 9. These are screenshots from the CUNY HTC/Response Rate map. The yellow outline in each screenshot delineates the community's border. The yellow highlighting on each map shows the extent of the local Update/Leave operation. The rate increases between June 12 and June 15 are highlighted in red ovals around the trendline graph to the left of each map.

FIGURE 6 Island Heights, NJ& (estimated population 1,457, based on the 2014-18 American Community Survey): response rate almost tripled from 19% on June 12 to 54% on June 15 (Click to view map online)

 

 

FIGURE 7 Bishop Hill, IL (population 114): response rate more than doubled from 30% to 75% (Click to view the map online)

 

FIGURE 8 Wyoming County, WV (population 21,711): response rate more than doubled from 15% to 33.7% (Click to view the map online)

 

FIGURE 9 Lincoln County, CO (population 5,548): response rate almost doubled from 31% to 51.4% (Click to view the map online)

 

Response rates also increased substantially in many tribal areas covered by the Update/Leave operation. Figure 10 shows the daily response rates from March 20 through June 18 for 73 of the largest tribal areas (populations of 1,000 or more) that had low response rates (below 30%) before June 12.

The chart shows that between June 12 and June 15, the response rates in almost a third (24) of these tribal areas increased substantially, between 5 and 22 percentage points. Response rates in more than half (38) of these areas increased by 1 point or more.

FIGURE 10

INTERNET FIRST VS INTERNET CHOICE: RATES INCREASING FOR TRACTS THAT RECEIVED BILINGUAL ENGLISH-SPANISH MAILING

In the first (initial) mailing of census materials in mid-March, about 80% of homes in the mail-out universe (95% of all residential addresses in the U.S.) received Internet First packets, which included a letter of invitation to respond on-line and a language assistance sheet with toll-free numbers to respond by phone. The remaining 20% of homes in the mail-out universe received Internet Choice packets, which included the same materials plus a paper questionnaire.

Beginning on April 8, the Census Bureau mailed paper questionnaires and reminder letters to non-responding households in mail-out areas. Therefore, at this point, all households that originally received either an Internet First or an Internet Choice mailing have received a paper questionnaire if they hadn't already self-responded by the time of the fourth, targeted mailing in April.

Our analysis below reveals important gains in tracts with populations that are predominantly people of color. (Note that this comparison omits tracts with 10% or more housing units covered by the Bureau's Update/Leave operation, where census field staff hand-deliver Internet Choice packets to households and update the master address list as they go. The Update/Leave operation was suspended on March 18 due to COVID-19 safety concerns. A phased restart of this operation began on May 6.)

Table 3 presents response rates for tracts where homes received Internet Choice packets compared with response rates for tracts where homes received Internet First packets, displayed for tracts based on which racial group or population of Hispanic origin is the predominant population based on the most recent American Community Survey 5-year estimates (2014-2018).

Table 4 shows the change in response rates for these tracts between June 4 and June 18. It reveals that tracts that are predominantly Black or Hispanic and where households received bilingual mailings in March from the Census Bureau had notable gains in self-response over that time period. Table 4 also shows that tracts that are predominantly Asian or Hispanic and received the Internet Choice mailing in March had larger response rate increases than other tracts between June 4 and June 18.

TABLE 3

NB: Most of the 225 census tracts nationwide where the plurality population is American Indian/Alaska Native (single race, non-Hispanic) are in areas covered by Update/Leave and are omitted from this table for now.

TABLE 4

RESPONSE RATE TRENDS ACROSS CENSUS TRACTS BY PLURALITY RACE/HISPANIC ORIGIN

We report on response rates below for communities whose populations have substantial concentrations of groups that historically have been undercounted or are at risk of being undercounted in 2020 due to predicted low self-response rates. Importantly, this does not indicate anything about the population in households that have responded. It only compares the response rates for census tracts that have certain population characteristics. All population data are based on estimates for the 2014-2018 period from the American Community Survey.

People of color

We focus our analysis for Weeks 12 and 13 on people who reported their race during the 2014-2018 American Community Survey as a single race but not of Hispanic origin, and separately for the population that reported Hispanic origin regardless of race.

We examined the latest response rates for tracts based on plurality population for each racial group and Hispanic origin. We cross-tabulated these tracts based on population size of the city in which they are located, as well as for governmental units (incorporated places) outside of metropolitan areas and for unincorporated places. The results are shown in Table 5. The analysis omits tracts with fewer than 100 householders, and tracts with 10% or more units covered by the Census Bureau's Update/Leave operation.

The main finding presented in Table 5 is that tract-level rates continue to be lowest for all types of tracts in the largest cities across the country, regardless of which race or ethnic (i.e. Hispanic origin) group represents the plurality of the population.

TABLE 5 Response rates as of June 18 for tracts with a plurality of each group, by city size

Despite the continued low rates in large cities, Table 6 shows that self-response rates have been increasing in tracts whose populations are predominantly Hispanic across all city sizes. (The small number of tracts whose population is plurality American Indian/Alaska Native or Native Hawaiian/Pacific Islander also have had notable increases.) In the nation's largest cities, response rates in tracts that are predominantly Black have increased the most: almost 1 percentage point, on average, between June 4 and June 18.

TABLE 6 Percentage point response rate increases for tracts with a plurality of each group, by city size

NEW ONLINE RESPONSE RATE METRIC TO HELP SUPPORT GOTC WORK DURING THE EXTENDED SELF-RESPONSE TIMEFRAME

We have added a new feature to the CUNY HTC/Response Rate to help census stakeholders measure the progress of local Get Out the Count efforts to boost 2020 Census participation: a dynamic metric to quantify how many households need to self-respond to meet each local community's final 2010 response rate goal.

There are still more than four months remaining in the extended self-response timeframe that the Census Bureau established in response to the COVID-19 crisis. Householders can self-respond through October 31, 2020, although the Nonresponse Follow-up (NRFU) (door-knocking) operation to count non-responding households in person will begin on August 11 and continue through the end of October. This means that households can take advantage of the self-response option for an additional three months compared to the original pre-COVID-19 census schedule.

The new HTC/Response Rate map feature is designed to provide stakeholder groups with a manageable and measurable metric at the local level to track progress during the remaining self-response timeframe. A goal of many groups and local public officials is to meet and possibly surpass their community's final 2010 self-response rate. Simply comparing that rate with the current self-response percentage may not provide tangible markers for local groups with limited resources to determine how many households they need to convince to respond in order to meet the 2010 goal.

Therefore, the HTC/Response Rate map now displays for each census tract the estimated number of households that would need to complete the census each day, on average, between now and July 31 (the formal end of the Self-Response Operation, before census takers begin knocking on doors) to meet the tract's 2010 response rate. The map also shows this information by city and county as the average number of households per day per tract that would need to fill out the census to meet the city or county's 2010 goal.

The following figures are examples of how this information is displayed via the online map. This link displays a tract in Chicago, IL (also displayed in Figure 11). The left-hand panel shows that if an average of 1.2 households per day fill out the census by July 31, the tract will meet its 2010 rate.

FIGURE 11 (Click to view the map online)

If a tract has already achieved its 2010 rate, we note that as well. Figure 12 displays a tract in Chicago that has equaled and surpassed its 2010 response rate.

FIGURE 12 (Click to view the map online)

Click to view the map online

Figure 13 displays the same information for Chicago overall, indicating that if only three households, on average, fill out the census form per tract (across each of the city's 793 tracts) per day between now and July 31, Chicago would meet its 2010 goal.

FIGURE 13 (Click to view the map online)

These calculations are based on data from the Census Bureau's "Address Count Listing Files," representing the number of residential addresses per census block as of October 2019 in the Bureau's Master Address File (MAF). The MAF is the universe of housing units that are contacted by the Census Bureau to respond to the 2020 Census. The October 2019 data do not represent the final count of addresses in the MAF, but they are the most recent publicly available count of residential addresses prior to the launch of the 2020 Census.

In order to determine how many households still need to respond to meet each community's 2010 goal, we multiply the percentage point difference between the current response rate and the 2010 rate by the number of residential addresses in each tract from the Address Count Listing files and divide that by the number of days remaining between the current date and July 31. For cities and counties, we also divide that result by the number of tracts in the city or county.

It is important to understand that these calculations provide approximations of the actual number of households that would need to respond. But these estimates can be a helpful tool for stakeholder groups that need to quantify in a manageable way the number of households that need to fill out the census in order to at least meet and hopefully surpass each community's 2010 response rate goal.

Prepared June 7, 2020

Highlighted topics:

• NYC's response rate continues to outpace U.S. increases;
• The latest response rates in states where Update/Leave operations resumed as of May 6: modest increases;
• Closer look at communities already meeting their final 2010 response rates;
• Update on Internet First vs Internet Choice gap; and
• Update on response rates in plurality “historically undercounted population” census tracts.

Download a presentation of the report below
Summary presentation of the report below (forthcoming)

---

NATIONAL RESPONSE RATE TRENDS

The nationwide 2020 Census self-response rate as of Thursday, June 4 was 60.6%. The 2020 Census response rate at the national level continues to increase, but more slowly than in prior weeks. The rate on June 4 was less than 1 point above the rate two weeks ago on May 21.

Nonetheless, the nation has now surpassed the Census Bureau's projected 60.5% self-response rate for 2020. The Bureau had projected achieving this rate on or before April 30, when the self-response operation would have been completed before the COVID-19 pandemic required adjustments to the schedule, and the door-knocking operation (Nonresponse Follow-up, or NRFU) would have begun in mid-May.

The response rate trend is shown in the chart in Figure 1 below.

FIGURE 1

State-by-state response rates

The following charts show the latest 2020 response rates by state and the percentage point change by state from March 20 (when the Census Bureau first reported response rates) through June 4.

Note the inclusion of Puerto Rico in the charts in Figures 2 and 3. Now that the Update/Leave operation has resumed in Puerto Rico, we will be closely tracking self-response rate updates in the Commonwealth. (The entire Commonwealth is covered by the Update/Leave operation.)

FIGURE 2 States ranked by June 4 census response rate (the final 2010 U.S. rate and the current U.S. 2020 rate are included for comparison).

FIGURE 3 States ordered left-to-right by lowest response rate to highest on June 4 (including the current national rate and final 2010 response rate for comparison). Brown bars represent the beginning response rate (March 20); the blue bars represent the increase between 3/20 and 6/4. Click to view larger image.

Despite the slowing rate of increase at the national level, there are again important examples of response rates improving at the local level (we reported similar local gains compared to the national rate increase as of May 21).

Figure 4 below shows the weekly response rate increases (in percentage points) for the U.S. (the black bars) compared with cities with populations of 1 million or more (bars of different color).

In the first week (response rates reported from March 20 through March 26), the U.S. overall had a larger response rate than each of these cities, and New York City had the lowest response rate. In weeks 2 through 5, the response rate increases on a weekly basis in most of these cities surpassed the U.S. weekly increases. In weeks 6 through 8, that pattern reversed, with the U.S. weekly rate increase greater than in most large cities. In Weeks 10 and 11, the weekly increases in several cities – including, notably, New York – began to outpace the U.S. weekly increase. Other cities such as San Antonio, TX, and Philadelphia, PA, also outpaced the U.S. weekly increase.

In the last two weeks, New York City's response rate increased twice as much as the U.S. increase. In New York, the rate increased from 49.6% on May 21 to 51.2% on June 7 – an increase of 1.6 percentage points. Nationwide in that period, the rate increased 0.8 points.

FIGURE 4 (click to view larger image)

Key areas of concern analyzed for Weeks 10 and 11

In addition to these overall response rate trends, we focus our analysis for Weeks 10 and 11 on the following issues:

1. The latest response rates in states where Update/Leave operations resumed on a phased basis starting May 6
2. Areas already meeting their final 2010 response rates
3. Update on Internet First vs Internet Choice response rate gap
4. Update on response rates in plurality "historically undercounted population" and "low self-response rate" census tracts

---

1. LATEST RESPONSE RATES IN STATES WHERE UPDATE/LEAVE OPERATIONS HAVE RESUMED

On May 6, the Census Bureau began a phased restart of the Update/Leave operation, which involves hand-delivering census packets to households in some rural and remote areas, all of Puerto Rico, and on Tribal lands. According to the Census Bureau, the phased restart will proceed as state, local, and Tribal health conditions and restrictions permit.

Operations were restarted on May 6 across 13 states. Nine more states were added starting May 13, and operations resumed in additional states – as well as in Puerto Rico and the District of Columbia – on May 20, May 27, and June 3. The Update/Leave operation covers 5 million housing units stateside in 2020, plus approximately 1.7 million units in Puerto Rico; however, most of these homes (about 95%) did not receive their packets before the Bureau suspended the effort soon after it began in mid-March due to COVID-19 restrictions.

The map in Figure 5 below highlights these states and the Area Census Offices (ACOs) within them that will oversee the hand-delivery of census packets. The Bureau is announcing weekly where operations will resume the following Wednesday; visit the Census Bureau's website each Friday afternoon for updates.

Note that in some states, the Update/Leave operation did not resume (or is not resuming) in all Area Census Offices at the same time. Also, the ACOs shown in grey are where hiring is resuming for future census field operations, but these ACOs do not include any housing units covered by Update/Leave. The CUNY HTC/Response Rate map includes a search feature that makes it easy to zoom to any Area Census Office on the map, view the status of the resumption of field operations, and visualize areas within the ACO where Update/Leave operations will or have restarted.

FIGURE 5

Now that it has been several weeks since Update/Leave operations resumed in 22 states (either on May 6 or May 13), we have examined response rates in tracts in these states where most housing units are covered by Update/Leave, compared with rates in tracts where most or all units received their census packets by mail. Our findings are discussed below.

States in "Week 1" group (where U/L operations resumed 5/6/20)

We compared rates on May 6 with rates four weeks later (June 3):

• Overall, there are only modest response rate increases for majority U/L tracts over this period.
• We continue to see similar patterns to changes noted as of May 21 (Weeks 8 and 9 report):
• Tracts with majority of housing units that received census packets by mail had greater rate increases than Update/Leave tracts.
• Majority U/L tracts tended to have greater increases for internet response than mail/phone response.

Tables 1 and 2 below show these patterns.

TABLE 1

TABLE 2

States in "Week 2" group (where U/L operations resumed 5/13/20)

We compared rates on May 13 with rates three weeks later (June 3):

• Overall, there are only modest response rate increases for majority U/L tracts over this period.
• But in several Week 2 states, majority U/L tracts had response rate increases on par with majority mail-out tracts.
• Similar to Week 1 states, majority U/L tracts tended to have greater increases for internet response than mail/phone response.

Tables 3 and 4 below show these patterns.

TABLE 3

TABLE 4

In evaluating the pace of self-response in these areas, keep in mind that the length of time it takes to hand-deliver census packets to all households within the Update/Leave operation in each ACO jurisdiction depends on workload, geographic dispersion of housing units, and census field worker productivity. Once packets have been left at the front door, residents might not be aware that the materials are there immediately, especially because some people are staying indoors more than usual due to the pandemic.

In addition, there have been reports that some rural Post Offices have closed temporarily; residents in Update/Leave areas might not have a way to mail back a paper questionnaire easily or quickly. The Census Bureau also has not added targeted advertising in Update/Leave areas to its communications campaign. In mid-June, the Bureau will mail a postcard to households in the Update/Leave operation that only receive mail at a post office box, letting them know that their census packet has been or soon will be delivered to the door of their home. All of these factors could affect the pace of self-response increases in census tracts where all or some households are covered by the Update/Leave operation.

2. POPULATION CHARACTERISTICS OF CENSUS TRACTS ALREADY MEETING THEIR 2010 RESPONSE RATES

Based on the response rates published on June 4, just over 12,000 census tracts have exceeded their 2010 rates, representing 14% of all census tracts nationwide. This is an increase from 9,300 tracts on May 21. The average response rate for tracts that surpassed their 2010 response rates as of June 4 is 71.1%, which is 10.5 points above the U.S. rate.

Just over 50 million people live in these census tracts, according to American Community Survey estimates for the 2014-18 5-year period. But the population characteristics for these tracts (also based on the 2014-18 ACS data) are different from the historically undercounted communities in which census stakeholders and civil rights groups have focused their 2020 Census outreach efforts.

Combined population characteristics for the 12,000 tracts that have surpassed their 2010 response rates are as follows:

• 68% White (non-Hispanic)
• 8% Black
• 13% Hispanic
• 7% Asian
• Housing is 70% owner-occupied
• 3% of households are limited English proficient (LEP)
• Population in poverty: 10%
• Roughly even distribution by educational attainment (people age 25+):
  • 31% high school diploma or less
  • 31% college (less than bachelor's degree)
  • 38% bachelor's degree or higher

3. UPDATE ON INTERNET FIRST vs INTERNET CHOICE RESPONSE RATE GAP

In the first (initial) mailing of census materials in mid-March, about 80% of homes in the mail-out universe (95% of all residential addresses in the U.S.) received Internet First packets, which included a letter of invitation to respond on-line and a language assistance sheet with toll-free numbers to respond by phone. The remaining 20% of homes in the mail-out universe received Internet Choice packets, which included the same materials plus a paper questionnaire.

Beginning on April 8, the Census Bureau mailed paper questionnaires and reminder letters to non-responding households in mail-out areas. At this point, all households that originally received either an Internet First or an Internet Choice mailing have received a paper questionnaire.

Our analysis below examines the ongoing response rate gap between Internet First and Internet Choice tracts. We find that the gap is still large, but it is decreasing.

On June 4, the average response rate was 65.3% across all tracts nationwide where homes received Internet First packets. This rate was almost 5 points above the June 4 U.S. rate of 60.6%. A month ago, on May 7, the average "Internet First" rate also was about 5 points above the U.S. rate at the time.

The average response rate on June 4 across all tracts nationwide where homes received Internet Choice packets was only 52.5%. This was 8 points below the June 4 U.S. rate. This gap, while troubling, reflects an improvement; a month ago, on May 7, the average "Internet Choice" response rate was 9 points below the U.S. rate.

The gap between the average response rate for Internet First and Internet Choice tracts also has improved – it is now just under 13 points. On May 7, it was 14 points. The gap is substantial, but it is decreasing, indicating that more homes in Internet Choice communities relative to Internet First communities are responding to the census.

(Note that this comparison omits tracts with 10% or more housing units covered by the Bureau's Update/Leave operation, where census field staff hand-deliver Internet Choice packets to households and update the master address list as they go. The Update/Leave operation was suspended on March 18 due to COVID-19 safety concerns. A phased restart of this operation began on May 6.)

This decrease in the gap between Internet First and Internet Choice tracts, while modest overall, is especially noticeable for tracts whose populations are predominantly people of color.

Table 5 below presents response rates for tracts where homes received Internet Choice packets compared with response rates for tracts where homes received Internet First packets, displayed for tracts based on which racial group or population of Hispanic origin is the predominant population.

TABLE 5

Table 6 below shows the change in response rates for these tracts between May 7 and June 4. The table highlights that rates for Internet Choice tracts have greater increases than Internet First for all tracts as categorized by plurality people of color.

Similarly, the table shows that rates are increasing the most for tracts with plurality populations of color that received bilingual (Spanish and English) Internet Choice census packets.

TABLE 6

4. UPDATED RESPONSE RATE TRENDS FOR CENSUS TRACTS WITH PLURALITY POPULATIONS OF HISTORICALLY UNDERCOUNTED GROUPS

We report on response rates below for communities whose populations have substantial concentrations of groups that historically have been undercounted or are at risk of being undercounted in 2020 due to predicted low self-response rates. Importantly, this does not indicate anything about the population in households that have responded. It only compares the response rates for census tracts that have certain population characteristics. All population data are based on estimates for the 2014-2018 period from the American Community Survey.

People of color

We focus our analysis on people who reported their race during the 2014-2018 American Community Survey as a single race but not of Hispanic origin, and separately for the population that reported Hispanic origin regardless of race.

We examined the latest response rates for tracts based on plurality population for each racial group and Hispanic origin. We cross-tabulated these tracts based on population size of the city in which they are located, as well as for governmental units (incorporated places) outside of metropolitan areas and for unincorporated places. The results are shown in Table 7. The analysis omits tracts with fewer than 100 householders, and tracts with 10% or more units covered by the Census Bureau's Update/Leave operation.

The main finding presented in Table 7 is that tract-level rates continue to be lowest for all types of tracts in the largest cities across the country.

TABLE 7 Response rates as of June 4 for tracts with a plurality of each group, by city size

(Note that in Table 7, 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.)

Despite the continued low rates in large cities, Table 8 below shows that response rates have been increasing in tracts whose populations are predominantly Black or Hispanic, and in the small number of tracts whose population is plurality American Indian/Alaska Native or Native Hawaiian/Pacific Islander (not including tracts that are covered by the Update/Leave operation). Also, response rates in tracts that are predominantly Black in the nation's largest cities have increased the most: 4.4 percentage points between May 7 and June 4.

TABLE 8 Percentage point response rate increases for tracts with a plurality of each group, by city size

Children under age 5

Children under age 5 represent the age cohort with the greatest net undercount in the decennial census. Although self-response rates do not tell us the extent of possible undercounting of any specific demographic group, nor do they indicate anything on their own regarding census accuracy, examining response rates in communities where children are at greatest risk of being undercounted can help census stakeholders target their outreach strategies and messaging to promote inclusion of young children in household census responses.

We examined response rates by tracts nationwide across 689 counties that the Population Reference Bureau (PRB) analyzed to determine tracts at greatest risk of an undercount of young children.

We found that tracts with "Very High Risk" of undercounting young children, according to PRB's analysis, continue to have an average response rate below the national level, at 54.9% as of June 4.

People with incomes below the poverty line

For this analysis, we focus on what are considered by poverty advocacy groups to be "High Poverty" tracts, in which more than 30% of residents for whom the poverty level has been determined have incomes below the poverty line.

The average response rate across all high poverty tracts nationwide as of June 4 is 46.9%, almost 14 points lower than the national average. This gap continues to widen; it was 13 points as of April 30 and 12 points as of April 9.

By contrast, tracts with less than 30% of the population in poverty had an average response rate as of June 4 of 64.3%, above the national average.

Educational attainment

We examined response rates by tract compared with education levels for the population age 25 and older. We found:

• In tracts where a plurality of people age 25+ have bachelor's degrees or higher, the average tract response rate is 68.6% (8 points above U.S. rate).
• In tracts where a plurality of people have a high school degree or less, average tract response rate is 58% (almost 3 points below U.S. rate).
• In the nation's largest cities (with populations of 1 million or more), response rates for both education groups are lower:
  • Tracts with a plurality of people holding bachelor's degree or higher have an average response rate of 58.2% (below the U.S. rate).
  • Tracts with a plurality of people holding a high school degree or less have an average response rate of 49.1% (9 points below higher-degree tracts and almost 12 points below U.S. rate).

Prepared May 25, 2020

Highlighted topics:

• Rates in some cities are outpacing U.S. increases again
• Restart of Update/Leave Operation: Is it making a difference (yet)?
• What do the lowest responding census tracts ("bottom 20%") look like?
• Response rates in tracts with a plurality of foreign-born population from selected countries

Download a presentation of the report below
Download a summary presentation of the report below

---

NATIONAL RESPONSE RATE TRENDS

The nationwide 2020 Census self-response rate as of Thursday, May 21 was 59.8%. The rate on May 21 was only 2.1 percentage points above the rate two weeks ago on May 7. Daily response rate increases have slowed; before May 7, the U.S. rate had been increasing an average of two points per week. If rate increases continue at this slower pace, the U.S. will not reach its final 2010 self-response rate of 66.5% until July (roughly a month later than when we last examined the national rates as of May 7).

Nonetheless, the extended self-response timeframe provides ample opportunity to continue to boost the rates and increase the pace of response rate increases.

The response rate trend is shown in the chart in Figure 1 below.

FIGURE 1

Uncharted self-response rate territory offers an unprecedented opportunity to boost rates

Just as in the 2010 Census, April 30 would have been the end of the 2020 self-response operation, although the option to self-respond remains available through the Nonresponse Follow-up (NRFU) door-knocking period. At this point in May 2010, the door-knocking phase had already begun; similarly, under the original 2020 Census plan, follow-up visits to unresponsive households would have started on May 13. Instead, the COVID-19 crisis prompted the Census Bureau to delay the NRFU operation until mid-August and extend the option for householders to submit their census responses on their own until October 31, 2020.

At this point and going forward, instead of looking back to 2010 for daily comparisons, we can look ahead to the now-extended response timeframe for 2020. The new timeframe provides an unprecedented opportunity to boost self-response rates and therefore reduce the universe of households that will need to be counted in-person — perhaps substantially. It also presents an important opportunity to change the historic pattern of geographic and demographic characteristics of the NRFU universe, by increasing self-response among historically undercounted households and population groups.

State-by-state response rates

The following graphs show the latest 2020 response rates by state and the percentage point change by state from March 20 (when the Census Bureau first reported response rates) through May 21.

Note the inclusion of Puerto Rico in the charts in Figures 2 and 3. Now that the Update/Leave operation is resuming in Puerto Rico, we will be closely tracking self-response rate updates in the Commonwealth. (The entire Commonwealth is covered by the Update/Leave operation.)

FIGURE 2 States ranked by May 21 census response rate (the final 2010 U.S. rate and the current U.S. 2020 rate are included for comparison).

FIGURE 3 States ordered left-to-right by lowest response rate to highest on May 21 (including the current national rate and final 2010 response rate for comparison). Brown bars represent the beginning response rate (March 20); the blue bars represent the increase between 3/20 and 5/21.

Despite the slowing rate of increases at the national level, there are several important examples of response rates improving at the local level.

Increases in response rates in several of the nation’s largest cities in recent weeks have surpassed the increases at the national level. Figure 4 below shows the weekly response rate increases (in percentage points) for the U.S. (the black bars) compared with cities with populations of 1 million or more (bars of different color).

In the first week, the U.S. overall had a larger response rate than each of these cities, and New York City (the largest city) had the lowest response rate. In weeks 2 through 5, the response rate increases on a weekly basis in most of these cities surpassed the U.S. weekly increases. In weeks 6 through 8, that pattern reversed, with the U.S. weekly rate increase greater than in most large cities. But in weeks 8 and 9, the weekly increases in several cities – notably New York – began to outpace the U.S. weekly increase. Other cities such as San Antonio, TX, and Philadelphia, PA, also outpaced the U.S. weekly increase.

For more details about New York City’s recent increases, see Weekly Report on 2020 Census Self-Response (Rates Issued on May 14) by the New York City Department of City Planning’s Population Division.

FIGURE 4

Also noteworthy: numerous communities across the country have exceeded their 2010 “goalpost” final response rates. This reinforces the message that the extended timeframe for 2020 Census self-response provides an opportunity for local communities and even regions and states to surpass their 2010 levels, despite the challenges to census outreach presented by the COVID-19 crisis. Get Out the Count organizers can view the final 2010 rates not as a ceiling, but as a floor from which their self-response rates can continue to rise.

Based on the response rates published on May 21, the following areas have exceeded their 2010 rates:

Michigan’s 9th congressional district (spanning Macomb and Oakland counties, just north of Detroit)

136 counties

• 31 in Kentucky (such as Jessamine)
• 25 in Virginia (such as Frederick)
• 3 in West Virginia; 1 in New Mexico

Just over 2,300 cities/incorporated places

• Over 100 each in Illinois, Kentucky, Michigan, N. Carolina, Ohio, Pennsylvania, & Wisconsin

Approximately 9,300 census tracts

• 11% of all tracts nationwide (such as this tract in the Bronx, NY)
• At least one in every state plus the District of Columbia

37 tribal areas

• Most of these 37 areas are small and are not part of the Update/Leave operation.
• Most tribal lands still have low response rates due, in significant part, to the suspension of the Update/Leave operation, meaning that most households in these areas have not received their census packets yet.

Key areas of concern analyzed for Weeks 8 and 9

In addition to these overall response rate trends, we focus our analysis for Weeks 8 and 9 on the following issues:

1. The latest response rates in states where Update/Leave operations have resumed
2. Census tracts with the lowest response rates nationwide, compared to areas with the highest rates
3. Exploration of response rates in tracts with a plurality of foreign-born population from selected countries

---

1. LATEST RESPONSE RATES IN STATES WHERE UPDATE/LEAVE OPERATIONS HAVE RESUMED

On May 6, the Census Bureau began a phased restart of the Update/Leave operation, which involves hand-delivering census packets to households in some rural and remote areas, all of Puerto Rico, and on many Tribal lands. According to the Census Bureau, the phased restart will proceed as state, local, and Tribal health conditions and restrictions permit.

Operations were restarted on May 6 across 13 states. Nine more states were added starting May 13, and on May 21 the bureau announced the resumption of operations in an additional 12 states for the following week. The Update/Leave operation covers 5 million housing units stateside in 2020, plus approximately 1.7 million units in Puerto Rico; however, most of these homes (about 95%) did not receive their packets before the bureau suspended the effort soon after it began in mid-March due to COVID-19 restrictions.

The map in Figure 5 below highlights these states and the Area Census Offices (ACOs) within them that will oversee the hand-delivery of census packets. The bureau is announcing weekly where operations will resume the following Wednesday; visit the Census Bureau’s website each Friday afternoon for updates.

Note that in some states, the Update/Leave operation is resuming only in certain Area Census Offices. Also, the ACOs shown in grey in Figure 5 are where hiring is resuming for future census field operations, but these ACOs do not include any housing units covered by Update/Leave. The CUNY HTC/Response Rate map includes a search feature that makes it easy to zoom to any Area Census Office on the map, view the status of the resumption of field operations, and visualize areas within the ACO where Update/Leave operations will restart.

FIGURE 5

Now that Update/Leave operations have resumed across 22 states starting either May 6 or May 13, we have examined response rates in tracts in these states where most housing units are covered by Update/Leave, compared with rates in tracts where units received their census packets by mail. Our findings are discussed below.

States in “Week 1” group (where U/L operations resumed 5/6/20)

Tracts in these 13 states with a majority of housing units that received census packets by mail had greater overall response rate increases than Update/Leave tracts:

• Mail-out tracts in most states had greater 2-week increases than the U.S. 2-week increase from May 7 to May 21
• Tracts in some states had greater increases for internet response; others had greater increases for mail response

Majority U/L tracts tended to have greater increases for internet response than mail response.

Tables 1 and 2 below show these patterns.

TABLE 1

TABLE 2

States in “Week 2” group (where U/L operations resumed 5/13/20)

Consistent with the general slower pace of response rate increases last week, the overall changes in these nine states are relatively small. Mail-out tracts in these states had slightly larger increases than majority U/L tracts.

In majority U/L tracts, the internet mode of response tends to have greater increases than mail/phone.

In majority Mail-out tracts, the increases are about the same for each mode.

Tables 3 and 4 below show these patterns.

TABLE 3

TABLE 4

In evaluating the pace of self-response in these areas, keep in mind that the length of time it takes to hand-deliver census packets to all households within the Update/Leave operation in each ACO jurisdiction depends on workload, geographic dispersion of housing units, and census field worker productivity. Once packets have been left at the front door, residents might not be aware that the materials are there immediately, especially because some people are staying indoors more than usual due to the pandemic.

In addition, there have been reports that some rural Post Offices have closed temporarily; residents in Update/Leave areas might not have a way to mail back a paper questionnaire easily or quickly. Finally, as of this analysis, the Census Bureau had not yet started new, targeted advertising in Update/Leave areas, which would help remind residents to look for census packets at their front doors. All of these factors could have affected the pace of self-response increases in census tracts where all or some households are covered by the Update/Leave operation.

2. TRACTS WITH THE LOWEST AND HIGHEST RESPONSE RATES: WHAT DO WE KNOW ABOUT THEIR POPULATIONS?

This analysis expands on our research from Week 4 and Week 5, focusing on census tracts in the “bottom 20%” of response rates nationwide, based on the latest 2020 Census rates. This enables us to understand the geographic patterns and concentrations of tracts that will eventually require the greatest share of the door-knocking (Nonresponse Follow-up, or NRFU) effort.

We use the latest data from the Census Bureau’s American Community Survey (ACS) (2014-2018 “5-year estimates”) to determine the characteristics of the overall population in these “bottom 20%” census tracts. This shows not only if historically undercounted or historically low self-response population groups are in these tracts, but if the tracts that eventually will require the greatest in-person enumeration effort are home to disproportionately high numbers of these groups.

Of course, this approach does not identify who is being counted as the self-response operation continues. It only identifies the demographic characteristics of census tracts overall. But it provides an indication of which population groups (and which communities) are at risk of being undercounted, and whether that risk is increasing or decreasing for historically undercounted groups if they continue to represent a large share of the population in the bottom 20% of tracts.

On May 21, when the Census Bureau began reporting response rates, tracts with response rates in the lowest fifth had a response rate of less than 50.4%. Based on ACS estimates for the 2014-2018 period, there are approximately 50 million people in these tracts, approximately 22.6 million total housing units, and approximately 18.2 million occupied housing units (i.e. households).

Some population characteristics of these tracts are as follows:

• Population in poverty
  • 12.5 million people (26% of population in bottom 20% of tracts)
• Language challenges
  • 1.7 million Limited English Proficiency households (9.2%)
• Non-Hispanic White population
  • 18.7 million (37.1%)
• Non-Hispanic Black population
  • 12.7 million (25.1%)
• Non-Hispanic Asian population
  • 2.3 million (4.5%)
• Non-Hispanic American Indian/Alaska Native population
  • 363,000 (0.7%)
• Non-Hispanic Native Hawaiian and Pacific Islander population
  • 75,000 (0.1%)
• Hispanic population
  • 15.1 million (30%)
• Households with no internet
  • 4.4 million (24.1%)
• Foreign-born population
  • 9.9 million (19.6%)
• People age 25+ with a bachelor’s degree or greater
  • 7 million (21.8% of population age 25+ in the bottom 20% of tracts)
• People age 25+ with a high school degree or less
  • 16.5 million (51.8% of population age 25+ in the bottom 20% of tracts)

(Note that this analysis omits tracts with 10% or more units in Update/Leave or fewer than 100 residents. This constraint affects the inclusion of tracts on many American Indian reservations, in particular, as the hand-delivery of census packets has not resumed on many Tribal lands due to COVID-19 restrictions.)

In order to view where these tracts are concentrated, or not, the CUNY HTC/Response Rate map includes the ability to display the bottom 20% of tracts across the country. As an example, Figure 6 below highlights the tracts with response rates in the bottom 20% in Detroit, MI. In this particular city, these low-response-rate tracts are especially concentrated.

FIGURE 6 (click to view on the online map)

 

 

In contrast to the population in the bottom 20% of census tracts on May 21, the population in the “top 20%” of tracts (those with the highest response rates) had a smaller share of historically undercounted or historically low self-response population groups. Tracts in the top 20% had response rates of 73.3% or more. Based on the 2014-18 ACS estimates, these tracts represent 67 million people (33% more than the bottom 20%), 26 million housing units, and 24.8 million households.

The same population groups in the top 20% of census tracts for May 21 are as follows:

• Poverty
  • 3.8 million people (8.8 million fewer than bottom 20% of tracts)
• Language challenges
  • 468K Limited English Proficiency households (72% fewer than the bottom 20%)
• Non-Hispanic White population
  • 52.5 million (more than 2.8x as many as in bottom 20%)
• Non-Hispanic Black population
  • 2.9 million (77% fewer)
• Non-Hispanic Asian population
  • 4.6 million (twice as many)
• Non-Hispanic American Indian/Alaska Native population
  • 165,000 (half as many)
• Non-Hispanic Native Hawaiian and Pacific Islander population
  • 69,500 (roughly the same)
• Hispanic population
  • 5.1 million (67% fewer)
• Households with no internet
  • 2.1 million (half as many)
• Foreign-born population
  • 6.8 million (31% fewer)
• People age 25+ with a bachelor’s degree or greater
  • 21.2 million (3x as many)
• People age 25+ with a high school degree or less
  • 12.3 million (25% fewer)

3. HISTORICALLY UNDERCOUNTED GROUPS

As in previous weeks, we report on response rates below for communities whose populations have substantial concentrations of selected groups that historically have been undercounted or are at risk of being undercounted in 2020. All population characteristics are based on estimates for the 2014-2018 period from the American Community Survey and do not tell us anything about the types of people or households that have responded so far to the 2020 Census.

Response rates across tracts for foreign-born population from selected countries

For Weeks 8 and 9, we focus our analysis on the foreign-born population by place of birth.

In particular, we examine response rates for tracts with a plurality of foreign-born population from the following countries (selected based on those that are the largest “source countries” for each community of color):

• Cuba, Dominican Republic, El Salvador, Mexico
• Ethiopia, Haiti, Jamaica, Nigeria
• China, Korea, Philippines, Vietnam

Response rates for tracts whose foreign-born populations are predominantly from Cuba, Dominican Republic, El Salvador, Mexico

Table 5 summarizes the average tract-level response rates according to size of city in which each tract is located. The table shows that:

• Rates in tracts in largest cities tend to be lowest (of all place sizes shown)
• Rates across cities of all sizes for tracts with foreign-born pluralities of people from the Dominican Republic are lower than the U.S. rate
• Higher rates outside cities/incorporated places, and even in small cities, for tracts with foreign-born plurality of people from Cuba, El Salvador, and Mexico

TABLE 5

Response rates for tracts whose foreign-born populations are predominantly from Ethiopia, Haiti, Jamaica, Nigeria

Table 6 summarizes the average tract-level response rates according to size of city in which each tract is located. The table shows that:

• Rates in tracts in largest cities tend to be lowest (of all place sizes shown)
• Higher rates outside cities/incorporated places, and even in small/medium cities, for tracts with foreign-born plurality of people from Ethiopia and Nigeria
• Rates across cities of all sizes for tracts with foreign-born pluralities of people from Jamaica and Haiti are lower than the U.S. rate

TABLE 6

Response rates for tracts whose foreign-born populations are predominantly from China, Korea, Philippines, Vietnam

Table 7 summarizes the average tract-level response rates according to size of city in which each tract is located. The table shows that:

• Rates in the largest cities tend to be lowest for tracts with foreign-born pluralities of people from China and Korea (but not the Philippines or Vietnam)
• Rates across cities of all sizes for tracts with foreign-born pluralities of people from Vietnam or the Philippines are higher than the U.S. rate
• Higher rates outside cities/incorporated places, and even in small/medium cities, for tracts with foreign-born plurality of people from all 4 countries

TABLE 7

---

Prepared May 10, 2020

Highlighted topics:

• More evidence of response rate boosts from 4th mailing/related outreach
• Demographics of Internet Choice vs Internet First
• Historically undercounted groups
• Update/Leave operations restarting in some states

Download a summary presentation of the report below
NB: Our next analysis will cover Weeks 8 and 9 combined, through May 21

---

National Response Rate Trends

The nationwide 2020 Census self-response rate as of Thursday, May 7 was 57.7%. The previous Friday (May 1) the Census Bureau reported a 1 percentage point increase from the day before (April 30). Since then the U.S. rate has increased more modestly. If the average daily increases for the past week continue, the U.S. will reach its 2010 self-response rate of 66.5% by June 6 (three days earlier than our previous projected milestone last week). The response rate trend is shown in the chart in Figure 1 below.

FIGURE 1

The long red line across the top of the chart shows the day-to-day response rates in 2010. Starting on May 1 after the self-response operation had ended, the 2010 trendline is generally flat, only increasing by 1.2 percentage points between May 1 and June 22 when the final 2010 self-response rate of 66.5% was achieved. Note that the Census Bureau did not publish daily response rates from 2010 at the state level.

Uncharted self-response rate territory

Just as in the 2010 Census, April 30 would have been the end of the 2020 self-response operation. At this point in May 2010, the door-knocking phase was about to begin; similarly, under the original 2020 Census plan, follow-up visits to unresponsive households would have started on May 13. Instead, the COVID-19 crisis prompted the Census Bureau to delay the Nonresponse Follow-up (NRFU) operation until mid-August and extend the option for householders to submit their census responses on their own until October 31, 2020.

At this point and going forward, instead of looking back to 2010 for daily comparisons, we can look ahead to the now-extended response timeframe for 2020. The new timeframe provides an unprecedented opportunity to substantially boost self-response rates and therefore substantially reduce the universe of households that will need to be counted in-person. It also presents an important opportunity to change the historic pattern of geographic and demographic characteristics of the nonresponse follow-up (NRFU) universe.

State-by-state response rates

The following graphs show the latest 2020 response rates by state and the percentage point change by state from March 20 (when the Census Bureau first reported response rates) until now.

FIGURE 2 States ranked by May 7 census response rate (the final 2010 U.S. rate and the current U.S. 2020 rate are included for comparison).

FIGURE 3 States ordered left-to-right by lowest response rate to highest on May 7 (including the current national rate and final 2010 response rate for comparison). Brown bars represent the beginning response rate (March 20); the blue bars represent the increase between 3/20 and 5/7.

Key areas of concern analyzed for Week 7

In addition to these overall response rate trends, we focus our Week 7 analysis on the following issues:

1. More evidence of response rate boosts from 4th mailing/related outreach
2. Demographics of Internet Choice vs Internet First census tracts
3. Historically undercounted groups
   • Week 7 focus: people of color by nativity (foreign-born or native-born)
4. Update/Leave operations restarting in some states

---

1. MORE EVIDENCE OF RESPONSE RATE BOOSTS FROM 4TH MAILING/RELATED OUTREACH

In the first (initial) mailing of census materials in mid-March, about 80% of homes in the mail-out universe (95% of all residential addresses in the U.S.) received Internet First packets, which included a letter of invitation to respond on-line and a language assistance sheet with toll-free numbers to respond by phone. The remaining 20% of homes in the mail-out universe received Internet Choice packets, which included the same materials plus a paper questionnaire.

Most 2020 census responses continue to be submitted via the online portal. According to the response rate data published on May 7, 47.2% of responses were submitted via the internet, while only 10.5% of responses were submitted by mail or phone. The mail response rate, however, increased 2 percentage points from the week before, from 8.5% to 10.5%, while the internet response increased only 1 percentage point.

Beginning on April 8, the Census Bureau mailed paper questionnaires and reminder letters to non-responding households in mail-out areas. Due to the COVID-19 crisis, the Census Bureau extended the timeframe of these mailings: paper questionnaires (and related reminder letters) were mailed through April 30, and reminder postcards are being mailed from April 27 to May 9. In addition, in April, the Bureau sent previously-planned reminder letters bearing unique ID numbers and telephone assistance information to “mailable addresses” in Update/Leave areas, despite most households not yet receiving their initial census packets due to COVID-19 related suspension of field operations.

On a local level, as we discussed in our Week 6 report, the increases from mail response during the targeted 4th mailing period have been especially noticeable in areas where households did not receive paper questionnaires in the initial mailing in March. These areas received an Internet First census packet, which provided a letter of invitation to respond online, as well as toll-free phone numbers to respond by phone.

Figure 4 presents an example of how responses by mail are increasing substantially in many communities where households did not initially receive a paper questionnaire but have since obtained a paper questionnaire via the 4th mailing. The image shows a screenshot from the CUNY HTC/Response Rate map for census tract 105.03 in Jasper County, Georgia. Each of this tract’s housing units received an Internet First mailing.

The trendline displays the daily response rate for this tract. On April 13, the tract’s overall response rate was 42.6%. Most of those responses (41.8%) were from the online portal and only 0.8% were via mail/phone.

By May 7, after the Census Bureau had mailed paper questionnaires to non-responding households in this tract, the internet response for this tract had increased several percentage points to 47.2%, but the mail/phone response increased from 0.8% to 11%.

FIGURE 4 (click to view this tract on the map)

 

By comparison, almost all homes in a tract in nearby Jefferson County, Georgia (tract 9603) received an Internet Choice mailing that included the paper questionnaire. Figure 5 below shows the 2020 response rate trendline from the CUNY HTC/Response Rate map for census tract 9603.

The trendline shows that most households in this tract that have responded have done so by mail or by phone. Between April 13 and May 7, however, the mail response rate in this tract changed only slightly, increasing by just over 2 percentage points, even though the non-responding households in this tract also received another paper form via the Census Bureau’s 4th mailing.

FIGURE 5 (click to view this tract on the map)

 

These two tracts are typical of many others across the country.

2. DEMOGRAPHICS OF INTERNET CHOICE vs INTERNET FIRST CENSUS TRACTS

The increases in mail/phone responses discussed above tend to be most noticeable in areas that received Internet First mailings in March. But response rates in communities that received Internet First mailings already tended to be high, while response rates in communities that received Internet Choice mailings are lagging.

On May 7, the average response rate was 62.6% across all tracts nationwide where homes had received Internet First packets. This rate was almost 5 points above the May 7 U.S. rate of 57.7%. A week earlier (on April 30) the average “Internet First” rate was 4 points above the U.S. rate at the time.

The average response rate on May 7 across all tracts nationwide where homes had received Internet Choice packets was only 48.5%. This was 9 points below the May 7 U.S. rate. A week earlier (on April 30) the average “Internet Choice” response rate was 8 points below the U.S. rate, and the prior week it was 6 points below.

The gap between the average response rate for Internet First and Internet Choice tracts also had grown to almost 14 points. A week earlier, the gap was 13 points, and the week prior it was 10 points.

(Note that this comparison omits tracts with 10% or more housing units covered by the Bureau’s Update/Leave operation, where census field staff hand-deliver Internet Choice packets to households and update the master address list as they go. The Update/Leave operation was suspended by March 20 due to COVID-19 safety concerns and state or local restrictions. A phased restart of this operation began on May 6.)

Table 1 below presents response rates for tracts where homes received Internet Choice packets compared with response rates for tracts where homes received Internet First packets, displayed for tracts based on which racial group or population of Hispanic origin was the predominant population.

In tracts across all racial groups (except for the small number of predominantly Native Hawaiian tracts) and in tracts of predominantly Hispanic origin, the average response rates for Internet Choice tracts were lower than for Internet First tracts. The gap is more pronounced for all groups between Internet First tracts that received English-only mailings and Internet Choice tracts with bilingual mailings.

TABLE 1

NB: Most of the 225 census tracts nationwide where the plurality population is American Indian/Alaska Native (single race, non-Hispanic) are in areas covered by Update/Leave and are omitted from this table.

3. HISTORICALLY UNDERCOUNTED GROUPS

We report on response rates below for communities whose populations have substantial concentrations of groups that historically have been undercounted or are at risk of being undercounted in 2020. All population characteristics are based on estimates for the 2014-2018 period from the American Community Survey and do not tell us anything about the types of people or households that have responded so far to the 2020 Census.

For Week 7, we focus our analysis on people who reported their race during the 2014-2018 American Community Survey as a single race but not of Hispanic origin, and separately for the population that reported Hispanic origin regardless of race.

Response rates across tracts by race/Hispanic origin and city size

We examine the latest response rates for tracts based on plurality population for each racial group and Hispanic origin. We cross-tabulate these tracts based on population size of the city in which they are located. The results are shown in Table 2. The analysis omits tracts with fewer than 100 householders, and tracts with 10% or more units covered by the Census Bureau’s Update/Leave operation.

The response rate patterns in Table 2 are similar to the patterns we have reported for previous weeks. In particular, tracts whose populations are predominantly Black or Hispanic tend to have the lowest response rates. Tracts whose populations are predominantly Asian also tend to have low rates in the largest cities (with populations of 1 million or more), but higher in smaller cities or incorporated places. Overall, response rates in the largest cities (with populations of 1 million or more) tend to be lowest for all groups.

(In our next analysis for response rates through May 21, we will examine the change in response rates between May 7 and May 21 for these categories.)

TABLE 2: Response rates as of May 7 for tracts with a plurality of each group, by city size.

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.

Response rates across tracts by race/ethnicity and nativity

We also examine the relationship between the latest response rates and share of population that is foreign-born compared with native born.

The Census Bureau cross-tabulates “nativity” (whether someone is born a U.S. citizen or not) with selected racial groups — people reporting Black alone (regardless of Hispanic origin), Asian alone (regardless of Hispanic origin), White non-Hispanic alone — and population of Hispanic origin.

Nationwide at the census tract level, for tracts with 100 householders or more and fewer than 10% of housing units covered by the Census Bureau’s Update/Leave operation, we found that:

• Response rates on May 7 tended to be lower in tracts with a greater share of Hispanic population, but that negative relationship was somewhat stronger for the foreign-born Hispanic population (the correlation statistic for native-born Hispanic population is -.294, while the statistic for foreign-born Hispanic population is -.312).
• We found a similar relationship between response rates and foreign-born Black population: rates tend to be lower as the share of Black population increases, and the negative relationship was strongest for foreign-born Black population.
• For the non-Hispanic White population, higher response rates are strongly correlated with higher shares of native-born non-Hispanic Whites (correlation statistic = .466).
• Response rates tend to increase as the share of Asian population increases, and there was only a slight correlation with either native-born or foreign-born Asian population.

However, these differences are starker when response rates across cities of different sizes, and for tracts outside cities, are examined.

Table 3 below compares the average tract-level response rates from May 7 by city size for tracts with a plurality of Hispanic population and where the foreign-born Hispanic population is either larger or smaller than the native-born Hispanic population.

Overall, the average response rate for tracts with a plurality of people who are foreign-born Hispanic is lower (48%) than tracts with a plurality of people who are native-born Hispanic (50%). In large cities with populations of 1 million or more, average response rates are lowest for both types of tracts, but are especially low (41.8%) for tracts with a plurality of people who are foreign-born Hispanic.

For tracts outside cities or incorporated places, however, tracts with a plurality of people who are foreign-born Hispanic are higher (54.5%) than tracts with a plurality of people who are native-born Hispanic (50.8%).

TABLE 3

Table 4 below compares the average tract-level response rates from May 7 by city size for tracts where the population is predominantly non-Hispanic Black, and where the foreign-born Black population is either larger or smaller than the native-born Black population.

There are relatively few tracts (only 182 nationwide) whose population is predominantly foreign-born Black (non-Hispanic). Overall, the average response rate is lower in these few tracts that are predominantly foreign-born Black (44.6%) than for tracts in which the population is predominantly native-born Black (47.6%).

Half of these tracts (99) are in the Bronx, Brooklyn, and Queens in New York City. In these predominantly foreign-born Black tracts in New York City, the average response rate is very low: 39.7%.

TABLE 4

Table 5 below compares the average tract-level response rates from May 7 by city size for tracts where the population is predominantly non-Hispanic Asian, and where the foreign-born Asian population is either larger or smaller than the native-born Asian population.

Nationwide, there are relatively few tracts (1,659) whose population is predominantly Asian, which is likely a function of the Asian population being more mixed with other race groups by tract than the Black, Hispanic, or White populations. The average response rate for these tracts are relatively high except for those tracts in the largest cities (with 1 million or more people), where the average tract response rate is 49.6%.

Of these predominantly Asian tracts, almost all are plurality foreign-born Asian. There are only 238 nationwide whose population is predominantly native-born Asian (non-Hispanic). Most of these tracts (201) are outside cities or incorporated places, and the average response rate in these tracts is slightly higher (64.7%) than the tracts outside cities that are predominantly foreign-born Asian.

TABLE 5

Table 6 below compares the average tract-level response rates from May 7 by city size for tracts where the population is predominantly non-Hispanic White, and where the foreign-born White non-Hispanic population is either larger or smaller than the native-born White non-Hispanic population.

Nationwide, there are relatively few tracts (only 126) whose population is predominantly foreign-born White (non-Hispanic). Overall in these tracts, the average response rate is lower (49%) than where the White non-Hispanic is predominantly native-born (62.7%).

Half of these tracts (66) are in New York City, in south Brooklyn and in central Queens. In these New York City tracts, the average response rate is low: 45.6%.

TABLE 6

For our analysis of the response rates during Weeks 8 and 9, we plan to further examine the relationship across tracts by country of origin of the foreign-born population, as well as for detailed ethnic groups.

4. UPDATE/LEAVE RESTARTING IN SOME STATES

The Census Bureau has announced that is resuming Update/Leave operations in selected Area Census Offices in selected states across the country. The map below shows the areas where Update/Leave will resume as of May 6 and separately as of May 13. The Bureau plans to announce additional areas where Update/Leave will begin as long as health and safety concerns can be met, and in coordination with state officials.

FIGURE 6

---

Prepared May 1, 2020

Highlighted topics:

• pace of daily change in rates for internet-only compared with mail/phone
• Internet Choice, Internet First (4th mailing with paper questionnaire boosting mail response, result most noticeable in Internet First areas, response rates in Internet Choice areas still lagging)
• demographic trends (rates in areas with concentrations of historically undercounted groups)

Download a presentation of the report below
Download a summary presentation of the report below

---

National Response Rate Trends

The nationwide 2020 Census self-response rate as of Thursday, April 30, was 54.6%. The U.S. rate increased at a faster pace in Week 6 than the week before, with an average daily increase of 0.3 points this week. This was driven by two days (April 24 and April 29) that each had a 0.6 percentage point increase over the day before.

The response rate trend is shown in the chart in Figure 1 below. The chart is annotated to show that as of April 8, the Census Bureau has been mailing paper questionnaires and reminder letters to non-responding households in mail-out areas. Due to the COVID-19 crisis, the Census Bureau has extended the timeframe of these mailings: paper questionnaires (and related reminder letters) were mailed through April 30, and reminder postcards are being mailed from April 27 to May 9. In addition, in April, the Bureau sent previously-planned reminder letters bearing the unique ID numbers and telephone assistance information to “mailable addresses” in Update/Leave areas, despite most households not yet receiving their initial census packets due to COVID-19 related suspension of field operations.

April 30 would have been the end of the 2020 self-response operation, just like that date marked the end of the 2010 self-response operation. Figure 1 below displays the daily self-response rates in 2010 at the national level compared with the national and state-by-state rates so far in 2020.

The long red line across the top of the chart shows the day-to-day response rates in 2010. Starting on May 1 after the self-response operation had ended, the 2010 trendline is generally flat, only increasing by 1.2 percentage points between May 1 and June 22 when the final 2010 self-response rate of 66.5% was achieved.

However, the individual lines in Figure 1 below representing state-by-state daily response rates in 2020 and the shorter red line noted with the arrow representing the 2020 U.S. daily rate, are all on an upward slope. Now that the 2020 self-response operation has been extended well beyond April 30, if the rates continue at their current pace, the nation will meet and then exceed its final 2010 self-response rate by June 9, two days earlier than we expected based on the prior week’s pace.

Note that the Census Bureau did not publish daily response rates from 2010 at the state level.

FIGURE 1

NB: Keep in mind the differences in census operations between 2010 and now when comparing response rate trends. Census operations were different in 2010, which may impact the difference in the national trend. For example, the 2010 Census was largely paper-based (the Bureau didn’t advertise the online option, which was used minimally); there were only two initial mailings (rather than staggered waves of initial mailings and multiple follow up mailings in 2020); there was a targeted replacement mailing in 2010; and of course, the 2010 Census did not face a pandemic that delayed key census operations and upended Get Out the Count (GOTC) campaigns that were based significantly on in-person outreach. Nonetheless, stakeholders will measure the success of the 2020 self-response operation by whether the nation, states, and localities meet and surpass their 2010 rates. Our map allows you to compare not only the 2020 rates with final 2010 “goalpost” rates, but also assess how your area is doing compared with nearby communities or similar areas across the state, and for larger regions such as counties, the state, and nation as a whole.

The following graphs show the latest 2020 response rates by state and the percentage point change by state from March 20 (when the Census Bureau first reported response rates) until now.

FIGURE 2 States ranked by April 30 census response rate (the final 2010 U.S. rate and the current U.S. 2020 rate are included for comparison).

FIGURE 3 States ordered left-to-right by lowest response rate to highest on April 30 (including the current national rate and final 2010 response rate for comparison). Brown bars represent the beginning response rate (March 20); the blue bars represent the increase between 3/20 and 4/30.

Key areas of concern analyzed for Week 6

In addition to the nationwide response rate trends, we focus our Week 6 analysis on the following issues:

1. Internet Choice, Internet First
   • 4th mailing with paper questionnaire boosting mail response
   • Result is most noticeable in Internet First areas
   • Response rates in Internet Choice areas, on average, still lagging behind Internet First
2. Demographic trends
   • Rates in areas with concentrations of historically undercounted groups
   • Some areas gained, others still face challenges

---

1. CHANGES IN INTERNET CHOICE AND INTERNET FIRST RESPONSE RATES

In the first (initial) mailing of census materials in mid-March, about 80% of homes in the mail-out universe (95% of all residential addresses in the U.S.) received Internet First packets, which included a letter of invitation to respond on-line and a language assistance sheet with toll-free numbers to respond by phone. The remaining 20% of homes in the mail-out universe received Internet Choice packets, which included the same materials plus a paper questionnaire.

Most 2020 census responses continue to be submitted via the online portal. According to the response rate data published on April 30, 46.1% of responses the day before were submitted via the internet, while only 8.5% of responses were submitted by mail or phone, for an overall response rate reported for April 30 of 54.6%.

But on a daily basis, combined responses by mail and phone have been increasing. (The Census Bureau does not separately report mail response and phone response. However, the Bureau noted recently that only 0.6% of responses have been submitted by phone.)

The blue line in the chart in Figure 4 below shows the daily change in response rates from internet responses, and the orange line represents daily change from mail/phone responses. From March 21 through April 11, the daily change from internet responses was greater than the daily change from mail/phone. (On April 2, reflecting the responses from "Census Day" on April 1, the daily increase from internet responses was much greater: a 2.7 percentage point increase from the internet, compared with only 0.2 percentage point increase from mail/phone.)

On April 12 and again on April 26, 27, and 29, the increase from mail response equaled the increase from internet response. And on April 19, April 22, and April 24 the increase in mail response surpassed the increase from internet response. These increases from mail (and to a much smaller extent, phone) response coincide with the Census Bureau mailing paper questionnaires to non-responding households (the 4th mailing, which arrived in mailboxes from April 8 – April 30).

FIGURE 4

Increases in mail response most noticeable in Internet First areas

On a local level, the increases from mail response have been especially noticeable in areas where households did not receive paper questionnaires from the Census Bureau in the initial mailing from the Bureau in March. These areas received an “Internet First” mailing, which provided a letter of invitation to respond online, as well as toll-free phone numbers to respond by phone.

Figure 5 presents an example of how responses by mail are increasing substantially in many communities where households did not initially receive a paper questionnaire from the Census Bureau. The image shows a screenshot from the CUNY HTC/Response Rate map for census tract 9634 in Seneca County, Ohio. Each of this tract’s housing units received an Internet First mailing.

The trendline from the map shown in the screenshot below displays the daily response rate for this tract. On April 17, the tract’s overall response rate was 56%. Most of those responses (54.8%) were from the online portal and only 1.2% were via mail/phone.

By April 30, after the Census Bureau had mailed paper questionnaires to non-responding households in this tract, the internet response for this tract had only increased slightly (from 54.8% to 56.5%) while the mail/phone response increased from 1.2% to 12.8%.

FIGURE 5 (click to view this tract on the map)

 

 

By comparison, the homes in another tract in Seneca County, OH (tract 9630) each received an Internet Choice mailing that included the paper questionnaire. Figure 6 below shows the 2020 response rate trendline from the CUNY HTC/Response Rate map for census tract 9630.

The trendline shows that most households in this tract that have responded have done so by mail or by phone. Between April 17 and April 30, however, the response rate in this tract changed only slightly, increasing by 1.3 points (from 56.2% to 57.5%). Although non-responding households in this tract have received a second paper questionnaire from the Census Bureau in the 4th targeted mailing, this apparently has not prompted these households to mail back the form.

FIGURE 6 (click to view this tract on the map)

 

 

The examples in Figures 5 and 6 above are typical of many tracts that are either “Internet First” or “Internet Choice” across the country.

Visualizing the increase in mailed responses from Internet First areas by county

This increase in mailed responses from Internet First areas is also noticeable at the county level.

Figures 7 and 8 below highlight counties in two states, Ohio and Kansas, as examples of this trend. In these two states, 76% of homes received Internet First packets in mid-March. The changes shown below are typical of states where most housing units received Internet First mailings.

Figure 7 shows the daily response trends for mail/phone only in Ohio counties where the share of homes that received the Internet First mailing exceeds the share that received the Internet Choice mailing (in other words, most homes in these counties did not receive a paper questionnaire in the initial mailing from the Census Bureau). This analysis also omits counties with an above-average share of homes that are covered by the Bureau’s Update/Leave operation.

Figure 7 shows that, beginning on or soon after April 17, after the Census Bureau had mailed paper questionnaires to non-responding households, the county-level rates for mail/phone response began to increase noticeably across all these counties.

FIGURE 7

Figure 8 shows the same types of counties in Kansas and reveals a similar pattern, although in Kansas, the mail/phone response rate increases tended to begin earlier, on or soon after April 11.

FIGURE 8

Internet Choice response rates still lagging behind Internet First

The increases in mail/phone responses discussed above tend to be most noticeable in areas that received Internet First mailings in March. But response rates in communities that received Internet First mailings already tend to be high, while response rates in communities that received Internet Choice mailings are lagging.

On April 30, the average response rate was 59.1% across all tracts nationwide where homes had received Internet First packets. This rate was more than 4 points above the April 30 U.S. rate of 54.6%. A week earlier (on April 23) the average “Internet First” rate was also 4 points above the U.S. rate at the time.

The average response rate on April 30 across all tracts nationwide where homes had received Internet Choice packets was only 46.3%. This was 8 points below the April 30 U.S. rate. A week earlier (on April 23) the average “Internet Choice” response rate was 6 points below the U.S. rate.

The gap between the average response rate for Internet First and Internet Choice tracts also had grown to almost 13 points. A week earlier, the gap was 10 points.

(Note that this comparison omits tracts with 10% or more housing units covered by the Bureau’s Update/Leave operation, where census field staff hand-deliver Internet Choice packets to households and update the master address list as they go. The Update/Leave operation was suspended by March 20 due to COVID-19 safety concerns and state or local restrictions.)

Table 1 below presents response rates for tracts where homes received Internet Choice packets compared with response rates for tracts where homes received Internet First packets, displayed by city size (and for tracts not in cities). The table also separately presents response rates for tracts where homes received bilingual packets (in Spanish and English) that were either Internet First or Internet Choice.

TABLE 1 (click to view a larger image)

2. DEMOGRAPHIC TRENDS

We report on response rates below for communities whose populations have substantial concentrations of groups that historically have been undercounted or are at risk of being undercounted in 2020. All population characteristics are based on estimates for the 2014-2018 period from the American Community Survey.

Children under age 5

Children under 5 represent the age group with the greatest net undercount in the decennial census. Although self-response rates do not tell us the extent of possible undercounting of any specific demographic group, nor do they indicate anything on their own regarding census accuracy, examining response rates in communities where children are at greatest risk of being undercounted can help census stakeholders target their outreach strategies and messaging to promote inclusion of young children in household census responses.

We examined response rates by tracts nationwide across 689 counties that were studied by the Population Reference Bureau (PRB), to determine tracts at greatest risk of an undercount of young children.

We found that tracts with “Very High Risk” of undercounting young children, according to PRB’s analysis, continue to have an average response rate below the national level, at 48.8% as of April 30.

People with incomes below the poverty line

For this analysis, we focus on what are considered by poverty advocacy groups to be "High poverty" tracts, in which more than 30% of residents for whom the poverty level has been determined have incomes below the poverty line.

The average response rate across all high poverty tracts nationwide as of April 30 is 41.3%, more than 13 points lower than the national average. This gap has widened; as of April 9, the average response rate across high poverty tracts was 12 points lower than the U.S. average.

By contrast, tracts with less than 30% of the population in poverty had an average response rate as of April 30 of 58%, above the national average.

The average response rate across these “low poverty” tracts is increasing almost twice as fast as the high poverty tracts. On April 9, “low poverty” tracts had an average response rate of 47%. This increased 11 percentage points by April 30, while the average rate for high poverty tracts only increased 6.8 percentage points.

People of color

We focus our analysis on people who reported their race during the 2014-2018 American Community Survey as a single race but not of Hispanic origin, and separately for the population that reported Hispanic origin regardless of race.

We examined the latest response rates for tracts based on plurality population for each racial group and Hispanic origin. We cross-tabulated these tracts based on population size of the city in which they are located. The results are shown in Table 2. The analysis omits tracts with fewer than 100 householders, and tracts with 10% or more units covered by the Census Bureau’s Update/Leave operation.

TABLE 2 Response rates as of April 30 for tracts with a plurality of each group, by city size

Response rates in tracts whose population is predominantly Black (non-Hispanic) or Hispanic are lower, on average, than tracts whose population is predominantly White (non-Hispanic) or Asian (non-Hispanic), and lower than the overall average response rate. This finding holds for census tracts in cities of any size and for tracts outside cities.

Average response rates for tracts whose population is predominantly Asian are higher in small and medium-size cities, and in tracts outside cities, than for tracts whose population is predominantly White. However, tracts that are predominantly Asian, Black, or Hispanic in the nation’s largest cities (with populations of 1 million or more) tend to have the lowest response rates.

When we last examined response rates for tracts nationwide on April 9 based on predominant population by race and Hispanic origin, we used a slightly different way to categorize tract population. At that time, we used American Community Survey estimates for people who reported their race “alone or in combination with other races” and regardless of Hispanic origin. That is a more expansive definition of who would be included in the categories of Black, Asian, or White. But people who are Hispanic can be included in “alone or in combination” categories such as White or Black, so we now analyze the data based on the “alone” race categories to eliminate this overlap.

After re-analyzing response rates from April 9 based on these non-overlapping categories, we calculated how average rates for tracts with a plurality of each group had changed between then and April 30. Table 2 presents the changes.

TABLE 3 Change in rates from April 9

Average response rates in tracts whose population is predominantly Asian increased the most in tracts outside cities and across tracts in cities smaller than 1 million people. Tracts whose population is predominantly Black had the smallest increases of all the groups in tracts across all city sizes and outside cities. However, tracts with predominantly Black populations had the largest response rate gains in large cities (with more than 1 million people) and in tracts outside cities.

Tracts whose population is predominantly Asian or Hispanic had the smallest gains in large cities of million or more.

Education level

We examined response rates by tract compared with education levels for the population age 25 and older. We found:

• in tracts where a plurality of people age 25+ have bachelor’s degrees or higher, the average tract response rate is 62.8% (8 points above the U.S. rate); and
• in tracts where a plurality of people have a high school degree or less, the average tract response rate is 51.5% (3 points below the U.S. rate).

In cities with populations of 1 million or more, rates for both education groups are lower:

• tracts with a plurality of people with bachelor’s degrees or higher have an average rate of 52.6% (almost at the U.S. rate); and
• tracts with a plurality of people with high school degrees or less have an average rate of 42.5% (10 points below the higher-educational attainment tracts and 12 points below the U.S. rate).

Prepared April 24, 2020

Highlighted topics:

• pace of daily change for internet-only compared with mail/phone
• city / metropolitan area / suburb / rural trends (two examples: Michigan/Detroit & Texas/Houston)
• demographic trends for tracts with lowest rates
• response rate gap between Internet Choice and Internet First tracts

Download a presentation of the report below
Download a summary presentation of the report below

---

National Response Rate Trends

The nationwide 2020 Census self-response rate as of Thursday, April 23, was 51.8%. There was a modest boost to the day-over-day increase on Wednesday April 22 (a 0.6 percentage point increase), but that dropped to a 0.2 percentage point increase the following day (Thursday April 23).

Earlier this month the Census Bureau indicated that the "current national response rate is on track for what was expected," per an Associated Press report. However, if the national daily response rates increase at the same pace as the last two weeks (roughly a 0.3 percentage point average daily increase), the nation will not achieve the Census Bureau's projected 2020 final self-response rate of 60.5% until May 22, 2020. If the daily rate increases continue at the same pace, the nation will not match its 2010 final self-response rate of 66.5% until June 11.

The response rate trend is shown in the chart in Figure 1 below. The chart is annotated to show that as of April 8, the Census Bureau has been mailing paper questionnaires and reminder letters to non-responding households in mail-out areas. Due to the COVID-19 crisis, the Census Bureau has extended the timeframe of these mailings: paper questionnaires (and related reminder letters) will continue to be mailed through April 30, and reminder postcards will be mailed from April 27 to May 9. In addition, the Bureau sent previously-planned reminder letters bearing the unique ID numbers and telephone assistance information to “mailable addresses” in Update/Leave areas, despite most households not yet receiving their initial census packets due to COVID-19 related suspension of field operations.

The chart visualizes the day-to-day response rates in 2010 (the long red line) compared with the rates so far in 2020 (the shorter red line noted with the arrow). The other lines represent state-by-state response rates in 2020. (The Census Bureau did not publish daily response rates from 2010 at the state level.)

FIGURE 1

NB: Keep in mind the differences in census operations between 2010 and now when comparing response rate trends. Census operations were different in 2010, which may impact the difference in the national trend. For example, the 2010 Census was largely paper-based (the Bureau didn’t advertise the online option, which was used minimally); there were only two initial mailings (rather than staggered waves of initial mailings and multiple follow up mailings in 2020); there was a targeted replacement mailing in 2010; and of course, the 2010 Census did not face a pandemic that delayed key census operations and upended Get Out the Count (GOTC) campaigns that were based significantly on in-person outreach. Nonetheless, stakeholders will measure the success of the 2020 self-response operation by whether the nation, states, and localities meet and surpass their 2010 rates. Our map allows you to compare not only the 2020 rates with final 2010 “goalpost” rates, but also assess how your area is doing compared with nearby communities or similar areas across the state, and for larger regions such as counties, the state, and nation as a whole.

The following graphs show the latest 2020 response rates by state and the percentage point change by state from March 20 (when the Census Bureau first reported response rates) until now.

FIGURE 2 States ranked by April 23 census response rate (the final 2010 U.S. rate and the current U.S. 2020 rate are included for comparison).

FIGURE 3 States ordered left-to-right by lowest response rate to highest on April 23 (including the current national rate and final 2010 response rate for comparison). Brown bars represent the beginning response rate (March 20); the blue bars represent the increase between 3/20 and 4/23.

The majority of responses continue to be submitted via the online portal. According to the response rate data published on April 23, 44.8% of responses the day before were submitted via the internet, while only 7% of responses were submitted by mail or phone, for an overall response rate reported for April 23 of 51.8%.

On a daily basis, the increase in response also has been driven by internet responses until recently, as shown in Figure 4 below. The blue line in the chart shows the daily change in response rate from internet responses, and the orange line represents daily change from mail/phone responses. From March 21 through April 11, the daily change from internet responses was greater than the daily change from mail/phone. (On April 2, reflecting the responses from "Census Day" on April 1, the daily increase from internet responses was much greater: a 2.7 percentage point increase from the internet, compared with only 0.2 percentage point increase from mail/phone.)

On two separate days (March 24 and April 13), the daily change in response rates from mail/phone was negative, as shown in the chart where the orange line drops below the horizontal gray line that indicates zero change from the day before. In other words, on March 24 and April 13 there were fewer mail/phone responses than the day before. On April 19, the daily increase from internet responses fell below the daily increase from mail/phone (i.e., there were fewer internet responses that day than mail/phone responses).

On April 22, the daily increase from mail/phone rose above internet responses; on that day there was a 0.6 percentage point increase in daily response, most of which (0.4 percentage points) was due to mail/phone response. The following day (April 23) the mail/phone increase fell back to 0 (there were the same number of mail/phone responses as the day before). Nonetheless, we have noticed in several tracts across the country a growing increase in mail/phone responses compared to internet responses, and we will continue to monitor this pattern in the coming week to detect if the Census Bureau's ongoing mailings of paper questionnaires and reminders will have helped to boost mail response.

FIGURE 4

Key areas of concern analyzed for Week 5

In addition to the nationwide response rate trends, we focus our Week 5 analysis on the following issues:

1. City/metropolitan area/suburb/rural trends
   • State-level rates can mask substantial regional/local differences
   • Two examples: Detroit and Houston
2. Demographic trends for tracts with lowest rates
   • Can these trends be changed by the time NRFU begins?
3. Response rate gap between Internet Choice and Internet First tracts
   • What types of communities are responding more slowly?

---

CITY / METROPOLITAN AREA / SUBURB / RURAL TRENDS

At the national level and state-by-state, the 2020 Census self-response rates are improving daily. Figure 5 below from the CUNY HTC 2020/Response Rate map shows that as of April 1, each state's response rate was below 50% (and the national rate was only 38.4%), while Figure 6 from the April 23 view of the map shows that the U.S. rate is above 50% and more than half the states each have rates above 50%.

FIGURE 5

FIGURE 6

But this national and state-level view masks local differences in response rates. Below we examine the differences in average response rates by tract for cities (population 50,000 or more) in metropolitan areas, suburban areas outside these cities, areas outside metropolitan regions, and tracts covered by Update/Leave to illustrate the uneven rates across the country. In particular, we focus on two examples: Michigan (with an emphasis on Detroit), and Texas (with an emphasis on Houston).

Michigan / Detroit

Michigan's response rate on April 23 was 58.2% — 6.4 points above the U.S. rate. The average rate for tracts in cities with at least 50,000 people in metropolitan areas in Michigan was lower: 55.2%. In tracts in metropolitan areas outside cities (suburbs, generally, and including incorporated places of fewer than 50,000 people), however, the rate was 64% — almost 6 points above the statewide rate and 12 points above the U.S. rate.

Outside metropolitan areas in Michigan, we examined average response rates for tracts with fewer than 10% of their units covered by Update/Leave. The average response rate in these tracts was 52.9% — below the statewide rate and below the average rate in metropolitan areas. Separately we examined response rates for "Update/Leave" tracts, almost all of which in Michigan are outside metropolitan areas. The average response rate for these tracts was very low: only 32.1%.

Figures 7 and 8 below display metropolitan areas across Michigan with red outlines that represent boundaries of metropolitan areas. In Figure 7, the map shows county-level response rates as of April 23. Figure 8 shows the patterns of the Census Bureau's 2020 contact strategies: the map highlights in yellow where housing units are covered by Update/Leave (in most areas across Michigan's upper peninsula, as well as some parts of northern Michigan).

FIGURE 7

FIGURE 8

Focusing on one of Michigan's city's, however, reveals that response rates in specific cities are often lower than the metropolitan average. In the Detroit-Warren-Dearborn metropolitan area, the average tract-level response rate is 55.2% (the Census Bureau does not publish response rates for metropolitan areas, so we identified tracts within metropolitan Core-Based Statistical Areas/CBSAs and averaged the response rates across all these tracts by metro CBSA). This average metropolitan rate is 3.4 points above the U.S. rate, and on par with response rates for Michigan's urban areas in general.

But within the metropolitan area, Wayne County's response rate is lower (53.9%), and Detroit's rate is substantially lower: 41.2%. (The city's rate is 10.6 points below the U.S. rate, and 17 points below Michigan's statewide rate.)

Detroit's population is predominantly Black, its households are mostly renters, and a third of the population has incomes below the poverty line — each of which are historically undercounted or low self-response populations. In addition, as shown in Figure 9 below, homes in most tracts in Detroit received the Census Bureau's "Internet Choice" mailing which included the paper questionnaire in the mailing as well as information on responding online or by phone. As our analysis has shown in prior weeks (and below for this week), Internet Choice tracts tend to have much lower response rates than Internet First tracts, where homes received a mailing with information only for responding online or by phone.

FIGURE 9 (click to view online map)

Texas / Houston

Our second example regarding metropolitan/suburban/rural response rate differences is Texas, with an emphasis on Houston. Response rates in Texas are lower than in Michigan, al they exhibit similar patterns across these regions.

The statewide response rate in Texas for April 23 was 47.3%. This is 4.5 points below the U.S. rate. The average rate for tracts in cities of 50,000 people or more in metropolitan areas in Texas is 48.5%, slightly higher than the statewide rate. In tracts in metropolitan areas outside cities (generally these are suburban tracts, but also include incorporated places of fewer than 50,000 people), the average response rate rises to 50.2%.

The average rate for "non-Update/Leave" tracts outside metropolitan areas is lower, at 42.4%. The average response rate for "Update/Leave" tracts (with 10% or more homes covered by Update/Leave) is the lowest across all these typologies: 26.7%.

Figures 10 and 11 below display metropolitan areas across Texas with red outlines representing metropolitan areas. In Figure 10, the map shows county-level response rates as of April 23. Figure 11 shows the patterns of the Census Bureau's 2020 contact strategies: the map highlights in yellow where housing units are covered by Update/Leave (only 3.5% of housing units in Texas are covered by Update/Leave, but these areas extend across much of the state).

FIGURE 10

FIGURE 11

In Houston, however, there are no Update/Leave areas. The "Greater Houston" metropolitan area has an average tract-level response rate of 45.4% (below the U.S. rate and below the Texas urban area response rate). Harris County within this metro area has a slightly higher response rate of 47%. Houston itself has a lower response rate: 44% (roughly 8 points below the U.S. rate and about 3 points below the Texas statewide rate).

As Figure 12 below shows, the tract-level response rates in and around Houston illustrate a pattern similar to many other urban areas nationwide: relatively low response rates within the city, higher rates in the immediately adjacent suburban tracts, but lower (sometimes lowest) response rates outside the metropolitan area.

FIGURE 12 (click to view online map)

RESPONSE RATES FOR THE BOTTOM 20% OF TRACTS

Who will be in the NRFU universe by August, and how will that have changed during the extended 2020 period for self-response?

This analysis expands on our research from Week 4, focusing on the “bottom 20%” of tracts based on the latest 2020 Census response rates. This enables us to understand the geographic patterns and concentrations of tracts that will eventually require the greatest share of NRFU effort.

This week, we slightly modified our methodology to determine the bottom 20% of tracts. Instead of omitting tracts that have any housing units covered by the Update/Leave operation, we only omitted tracts that have 10% or more housing units in Update/Leave. This enables us to examine more tracts, but still separate the tracts where suspension of the Update/Leave operation is causing very low response rates.

We use the latest data from the Census Bureau’s American Community Survey (ACS) (2014-2018 “5-year estimates”) to determine the characteristics of the overall population in these “bottom 20%” census tracts. This shows not only if historically undercounted population groups are in these tracts, but if the tracts that eventually will require the greatest Nonresponse Follow-up effort are home to disproportionately high numbers of these groups.

Of course, this approach does not identify who is being counted as the self-response operation continues. It only identifies the demographic characteristics of census tracts overall. But it provides an indication of which population groups (and which communities) are at risk of being undercounted, and whether that risk is increasing or decreasing for historically undercounted groups if they continue to represent a large share of the population of the bottom 20% of tracts.

Comparisons from Week 1 to Week 4, and then Week 4 to Week 5

On March 20, when the Census Bureau began reporting response rates, tracts with response rates in the lowest fifth had a response rate of less than 8.6% (approximately 15,000 census tracts). Based on ACS estimates for the 2014-2018 period (using our modified approach regarding U/L tracts), there are approximately 59 million people in these tracts, approximately 23.6 million total housing units, and approximately 20.8 million occupied housing units (i.e., households).

Some population characteristics of these tracts are as follows:

• 8.7 million people (~15% of population in bottom 20% of tracts) had incomes below the poverty line
• Language challenges: 1.3 million households (6.4%) were "limited English proficiency" (LEP)
• Non-Hispanic White population: 31.4 million (53.2%)
• Non-Hispanic Black population: 8.8 million (15%)
• Hispanic population: 13.8 million (23.4%)
• Households with no internet: 3.1 million (15%)

On April 16, tracts with response rates in the lowest fifth had rates less than 41.4%. This also represents approximately 15,000 tracts, but they are a different set of tracts than on March 20. Based on ACS estimates for the 2014-2018 period, the universe of population and housing units in these tracts had changed:

• overall population decreased about 14% to 50.7 million people;
• the number of housing units decreased to 22.4 million; and
• the number of occupied housing units (households) decreased to 18.2 million.

However, the population groups listed above increased their concentration in the bottom 20% of census tracts between March 20 and April 16:

• Population in poverty increased to 12.3 million (from 15% to 25% of population in bottom 20% of tracts)
• Language challenges: LEP households increased to 1.8 million households (9.9% of the population in the bottom 20% of tracts as of April 16)
• Non-Hispanic White population: fell to 19 million (37.6%)
• Non-Hispanic Black population: grew to 11.2 million (22%)
• Hispanic population: grew to 16.2 million (32%)
• Households with no internet: grew to 4.2 million (23.4%)

By April 23, because response rates overall had increased, tracts with response rates in the lowest fifth had rates less than 43.1%. Based on ACS estimates for the 2014-2018 period, the universe of population and housing units in these tracts had changed, but not substantially since the prior week. Overall population increased slightly to 50.9 million people, the number of housing units increased slightly to 22.6 million, and the number of occupied housing units (households) increased to 18.3 million.

The population groups listed above in the bottom 20% of census tracts also had only modest changes between April 16 and April 23:

• Population in poverty: slight increase (12.4 million households, or 25.4%)
• Language challenges: slight decrease in LEP households to 1.73 million (9.5%)
• Non-Hispanic White population: grew slightly to 19.3 million (37.8%)
• Non-Hispanic Black population: grew to 11.9 million (23%)
• Hispanic population: decreased to 15.7 million (31%)
• Households with no internet: increased slightly to 4.4 million (23.9%)

Week 5 comparisons between the Bottom 20% and Top 20% of census tracts

In contrast to the population in the bottom 20% of census tracts on April 23, the population in the top 20% of tracts (those with the highest response rates) had a smaller share of historically undercounted population groups. Tracts in the top 20% had response rates of 64.1% or more. Based on the 2014-18 ACS estimates, these tracts represent 66.5 million people (30% more than the bottom 20%), 25.8 million housing units, and 24.5 million households.

The population groups listed above in the top 20% of census tracts for April 23 are as follows:

• Population in poverty: 4 million people (8.5 million fewer than bottom 20%)
• Language challenges: 500,000 LEP households (70% fewer than the bottom 20)
• Non-Hispanic White population: 50.7 million (more than 1.5x as many as in the bottom 20)
• Non-Hispanic Black population: 3.1 million (74% fewer)
• Hispanic population: 5.6 million (65% fewer)
• Households with no internet: 2 million (53% fewer)

THE DEMOGRAPHICS OF INTERNET CHOICE TRACTS

Analysis: response rates for Internet Choice tracts are relatively low. What population groups are in these tracts?

About 80% of homes that received a mailed invitation from the Census Bureau for the 2020 Census (95% of all addresses) received Internet First packets (a letter of invitation to respond on-line, which also included toll-free phone numbers to call in the responses). The remaining 20% of housing units across the country received Internet Choice packets (these included the same materials as the Internet First mailing plus a paper questionnaire).

Response rates for Internet Choice tracts are increasing slowly compared with Internet First tracts. As of April 23, the response rate gap between the two had grown to more than 10 points. This week, the average response rate for Internet Choice tracts is more than 6 points below the U.S. rate; Internet First tracts have an average rate 4 points above the U.S. rate.

More people live in Internet First tracts (an estimated 244 million) compared with Internet Choice tracts (~59 million). But, “Choice” tracts tend to have:

• a greater share of population in poverty (24% vs 11%);
• a greater share of people of color. For example, non-Hispanic Black population is 26% of “Choice” tracts vs 9.3% of “First” tracts, and Hispanic population comprises 25% of “Choice” tracts vs 16.6% of “First” tracts; and
• a smaller share of non-Hispanic White population (44.4% in “Choice” tracts vs 64.3% in “First” tracts).

Internet First tracts also have more households (almost 91 million compared with 21.6 million). But Internet Choice tracts tend to have:

• a greater share of renter households (44% of “Choice” households vs 35% of “First” households);
• a greater share of “limited English proficiency” households (6.4% vs 4.1%); and
• a greater share of households with no home internet subscriptions (28% compared with 12.3%).

Prepared April 17, 2020

Highlighted topics:

• using HTC/Response Rate map to identify uneven response rates locally & regionally;
• new extended census timeframe (and new approach to assessing self-response/nonresponse follow-up implications for historically undercounted populations);
• Update/Leave and internet access; and
• Internet First compared with Internet Choice census tracts.

Download a presentation of the report below
Download a summary presentation of the report below

---

National Response Rate Trends

The nationwide 2020 Census self-response rate as of Thursday, April 16, was 49.4%. In the past week, the U.S. rate increased by just over 2 points. By comparison, in the second week of response rate reporting (March 27 through April 2) the U.S. rate rose by an average of 1.5 points per day.

The trend is shown in the chart in Figure 1 below. The chart is annotated to show that as of April 8, the Census Bureau has been mailing paper questionnaires and reminder letters to non-responding households. Due to the COVID-19 crisis, the Census Bureau has extended the timeframe of these mailings: paper questionnaires (and related reminder letters) will continue to be mailed through April 30, and reminder postcards will be mailed from April 27 to May 9.

The chart visualizes the day-to-day response rates in 2010 (the long red line) compared with the rates so far in 2020 (the shorter red line noted with the arrow). The other lines represent state-by-state response rates in 2020. (The Census Bureau did not publish daily response rates from 2010 at the state level.)

FIGURE 1

NB: Keep in mind the differences in census operations between 2010 and now when comparing response rate trends. Census operations were different in 2010, which may impact the difference in the national trend. For example, the 2010 Census was largely paper-based (the Bureau didn’t advertise the online option, which was used minimally); there were only two initial mailings (rather than staggered waves of initial mailings and multiple follow up mailings in 2020); there was a targeted replacement mailing in 2010; and of course, the 2010 Census did not face a pandemic that delayed key census operations and upended Get Out the Count (GOTC) campaigns that were based significantly on in-person outreach. Nonetheless, stakeholders will measure the success of the 2020 self-response operation by whether the nation, states, and localities meet and surpass their 2010 rates. Our map allows you to compare not only the 2020 rates with final 2010 “goalpost” rates, but also assess how your area is doing compared with nearby communities or similar areas across the state, and for larger regions such as counties, the state, and nation as a whole.

The following graphs show the latest 2020 response rates by state and the percentage point change by state from March 20 (when the Census Bureau first reported response rates) until now.

FIGURE 2 States ranked by April 16 census response rate (the final 2010 U.S. rate and the current U.S. 2020 rate are included for comparison).

FIGURE 3 States ordered left-to-right by lowest response rate to highest on April 16 (including the current national rate and final 2010 response rate for comparison). Brown bars represent the beginning response rate (March 20); the blue bars represent the increase between 3/20 and 4/16.

The pace of change in the response rate has been uneven on a state-by-state basis. The U.S. increased in Week 4 on average by 0.33 percentage points per day (Florida and Illinois also had the same daily increase). West Virginia had the fastest increase, of 0.5 points per day. The three states following West Virginia were Iowa, California, and Nebraska, tied at 0.44 points per day. The slowest increase was in Wisconsin, whose response rate rose on average by only 0.21 points per day during Week 4.

You can visualize the pace of change for any geographic location on our HTC 2020/Response Rate map. The following examples highlight a new feature added to the map: a trendline chart plotting day-to-day response rates for each of the 84,000 census tracts, roughly 34,000 cities and towns, 3,100 counties, 50 states and the District of Columbia, and the hundreds of tribal areas and congressional districts on the map (described further here).

In West Virginia, for example, Figure 4 below shows how the statewide response rate has increased from 9.8% on March 20 to 37.8% on April 16. The trendline is color-coded to match the legend on the right side of the map. The green line just below the color-coded trendline represents the rate based on internet responses. In West Virginia, where almost a third of the state’s housing units received the “Internet Choice” mailing — which included a paper questionnaire, the internet-only response rate of 28.7% on April 16 is almost 10 points lower than the overall statewide rate.

Figure 4 also highlights the one-day boost in response rates that likely reflects “Census Day” outreach activities on April 1.

FIGURE 4

In Figure 5 below, the map is zoomed in on West Virginia, and McDowell County is highlighted, along the Virginia border. In McDowell County, almost all (99.6%) of housing units are part of the now-suspended Update/Leave operation, and the county’s overall self-response rate is only 3.7%. Most (if not all) homes in the county haven’t yet received their initial census packets.

FIGURE 5

Figure 6 shows another county in West Virginia with a different response rate pattern: Wood County along the Ohio border. In Wood County, all housing units received a mailed invitation from the Census Bureau. The response rate for Wood County is much higher than McDowell County: 52%.

FIGURE 6

In Figure 7, the map is zoomed in and highlights Tract 108 in Wood County. In this tract, an estimated 9 homes out of more than 1,400 were to have received a hand-delivered census packet; the rest received an “Internet Choice” mailing. Consistent with most homes receiving a paper questionnaire in the mailing, most of the responses have been by mail or phone and fewer have been online (the internet-only response rate is just 22%).

Nonetheless, the tract’s 58% response rate is relatively high. Below in this analysis we show that nationwide, tracts that received Internet Choice mailings tend to have response rates lower than both the U.S. response rate and for tracts receiving Internet First mailings. This tract is an exception.

FIGURE 7

Key areas of concern analyzed for Week 4

In addition to the nationwide response rate trends, we focus our Week 4 analysis on the following issues:

1. New extended census timeframe (and a new approach to assessing self-response/nonresponse follow-up implications for historically undercounted populations)
2. Update/Leave and internet access
3. Internet First compared with Internet Choice census tracts

---

1. EXTENDED CENSUS 2020 TIMEFRAME / NEW ANALYSIS FOCUSED ON HISTORICALLY UNDERCOUNTED POPULATIONS IN TRACTS WITH LOWEST RESPONSE RATES

The Census Bureau’s original plan for 2020 was to publish response rates seven days per week from March 20 through May 31. The 2020 Census Nonresponse Follow-up (NRFU) operation was to have started in mid-May. Householders could still self-respond through July 31, so the Bureau planned to continue publishing response rates until then, but only on weekdays starting June 1.

Due to the COVID-19 crisis, after modest initial changes to the census schedule, the Bureau on April 13 announced significant adjustments to many of its operations. Now, householders will be able to self-respond through October 31, 2020 – an additional three months of self-response opportunity. NRFU will begin on August 11 (three months after it was originally scheduled to being) and will continue through the end of October.

In prior censuses, self-response rates typically increase only marginally once the formal self-response operation ends and NRFU begins. In 2010, for example, the U.S. response rate on April 30, 2010 (the end of the self-response operation) was 65.3%. Some households still responded on their own after NRFU began, but the nationwide response rate increased only slightly more than a percentage point during this time, peaking at 66.5% on June 22 (this represents an average daily increase from May 1 to June 22 of only .02 percentage points per day). See the daily 2010 response rates in Appendix H in this 2012 Census Bureau memo. [PDF]

Nonetheless, we have no historic context for how self-response rates will change over the course of such a longer timeframe.

Who will be in the NRFU universe by August, and how will that have changed during the extended 2020 period for self-response?

When our team at the CUNY Graduate Center was asked to develop an online map leading up to the 2020 Census to help stakeholders focus their education and outreach on the hardest to count communities and populations, the main question was how to define “hard-to-count.” In consultation with several census experts, we used an approach that identified the census tracts nationwide with the lowest shares of households that filled out the census questionnaire on their own during the last decennial census. These tracts had the most homes that needed to be counted in-person through NRFU. The NRFU door-knocking operation is challenging, expensive, and adds substantially to the risk that people will be missed or counted inaccurately. In other words, areas with the greatest share of homes that needed to be counted in-person were considered the hardest to count.

In order to identify tracts with the lowest shares of self-responding households, we used a threshold of the tracts in the bottom 20% of self-response rates nationwide. These tracts were then color-coded on the HTC 2020 map in orange-to-dark red as the “hardest to count” tracts.

We propose to use a similar approach during the 2020 Census self-response operation, to identify the census tracts that will have the greatest share of homes to be counted through NRFU starting in August 2020. But we will enhance this approach in two ways:

• We can determine the “bottom 20%” of tracts each day during the self-response operation based on the latest 2020 Census response rates. This enables us to understand not only where the tracts are located that will eventually require the greatest share of NRFU effort. But we can identify if the geographic concentrations of these tracts are changing, or if they remain concentrated in certain parts of the country.
• We can use the latest data from the Census Bureau’s American Community Survey (ACS) (2014-2018 “5-year estimates”) to determine the characteristics of the overall population in these “bottom 20%” census tracts. This will show not only if historically undercounted population groups are in these tracts, but if these groups are becoming more concentrated in the tracts with lowest self-response rates between now and when NRFU begins.

Of course, this approach will not identify who is being counted as the self-response operation continues. It only identifies the demographic characteristics of census tracts overall. But it provides an indication of which population groups (and which communities) are at risk of being undercounted, and whether that risk is increasing or decreasing for historically undercounted groups if they continue to occupy a large share of the population of the bottom 20% of tracts.

Initial results

In order to determine the threshold U.S. response rate for the tracts in the bottom 20% of response rates, we use the following assumptions:

• We are omitting tracts that include homes covered by the Update/Leave (U/L) operation. Response rates so far for U/L areas are low because most homes in those areas have not yet received census packets.
• We are also omitting tracts with very few people (any tract with fewer than 100 people as estimated by the American Community Survey is excluded from our analysis).

This provides a universe of approximately 65,000 tracts with more than 100 people each and with no U/L units. Twenty percent of this universe is approximately 13,000 tracts.

On March 20, when the Census Bureau began reporting response rates, tracts with response rates in the lowest fifth had a response rate of less than 8.6%. Based on ACS estimates for the 2014-2018 period, there are approximately 51.5 million people in these tracts, approximately 20.3 million total housing units, and approximately 18.1 million occupied housing units (i.e., “households”).

On April 16, tracts with response rates in the lowest fifth had rates less than 41.4%. This also represents approximately 13,000 tracts, but they are different tracts than on March 20. Based on ACS estimates for the 2014-2018 period, the universe of population and housing units in these tracts had changed:

• overall population decreased about 14% to 44 million people;
• the number of housing units decreased about 7% to 18.9 million; and
• the number of occupied housing units (households) decreased about 13% to 15.7 million.

However, certain population groups increased their concentration in the bottom 20% of census tracts between March 20 and April 16. The following examples illustrate this pattern. A detailed table is provided at the end of this section.

Population in poverty

Based on the March 20 threshold response rate of 8.6%, there were approximately 50 million people in those 13,000 tracts for whom the poverty rate had been determined. Of these, just over 7.5 million — approximately 15% of 50 million — had incomes below the poverty line.

By April 16, the population in the “bottom 20% of tracts” (the 13,000 tracts with a response rate of less than 41.4%) for whom poverty had been determined had decreased by 16% to about 42 million people. But people with incomes below the poverty line in those tracts had increased to almost 11 million (a 45% increase). The share of the population in poverty in the lowest response tracts grew by 11 percentage points, from almost 15% to almost 26%. In other words, a greater share of the population in the bottom 20% of tracts between March 20 and April 16 was living in poverty.

Households with “limited English proficiency”

Based on the March 20 threshold response rate of 8.6%, there were approximately 1.2 million households — about 6.7% of households — considered by the Census Bureau to have “limited English proficiency” (LEP) in the bottom 20%.

Although the estimated number of households in the bottom 20% of tracts decreased from March 20 to April 16 by about 13% (from approximately 18.1 million to 15.7 million households), the estimated number of LEP households increased by almost 500,000 during that time to 1.7 million households in the bottom 20% of tracts — an increase of more than 41%.

As a result, language challenges in the bottom fifth of tracts with the lowest response rates became greater. The share of LEP households in this group of tracts grew by 4 percentage points between March 20 and April 16.

People of color

We focus our initial analysis on two groups of people of color: those who reported their race during the 2014-2018 American Community Survey as Black but not of Hispanic origin, and separately for the population that reported Hispanic origin regardless of race.

By comparison, we also examine the population who reported their race during the 2014-2018 ACS as White but not of Hispanic origin.

Based on the March 20 threshold response rate of 8.6%, there were approximately 8 million people in the bottom 20% of tracts who were non-Hispanic Black. By April 16, the estimated non-Hispanic Black population in the bottom 20% of tracts increased by almost 33% to 10.5 million. Because the overall population of the bottom 20% of tracts declined between March 20 and April 16, the non-Hispanic Black population share increased, growing by just over 8 points (from just over 15% to almost 24%).

Based on the March 20 threshold response rate of 8.6%, there were approximately 12 million people in bottom 20% of tracts of Hispanic origin. By April 16, the estimated Hispanic population in the bottom 20% of tracts increased by almost 24% to approximately 15 million. Because the overall population in the bottom 20% of tracts declined between March 20 and April 16, the Hispanic population grew in concentration, increasing by more than 10 points (from just over 23% to almost 34%).

By comparison, the estimated non-Hispanic White population in the bottom 20% of tracts decreased overall and declined as a share of the population in the low-response-rate tracts.

Based on the March 20 threshold self-response rate, there were almost 27 million non-Hispanic Whites in the bottom 20% of tracts. By April 16, the White population in these tracts fell by about 12 million people (a 45% reduction). The concentration of non-Hispanic Whites as a share of population in the bottom 20% of tracts therefore decreased by more than 18 points (from just over 52% to almost 34%).

Table 1 below presents the data described above. It also includes statistics for other population groups historically at risk of being undercounted, such as renters, people in crowded housing, and single parent households. It also presents data for people who may be at risk of being undercounted in 2020 due to greater automation of the census process, such as households with no home internet subscriptions and householders age 65 or older. Click to enlarge.

TABLE 1

Based on feedback from census stakeholders on the value of analyzing population characteristics for tracts with the lowest self-response rates, we can build on this analysis in future weeks, such as mapping the bottom 20% of tracts to visualize if the geographic patterns are changing day-by-day or week-by-week. We can also examine other population groups, if requested.

2. INTERNET ACCESS IN UPDATE/LEAVE TRACTS

Analysis: are there differences across states in tract-level home internet access for areas with no units in Update/Leave, compared with areas that have at least some U/L coverage?

We found that in 44 states, tracts with at least some U/L areas had a greater share of homes without internet subscriptions, compared to tracts with no U/L areas.

Virginia had the greatest difference in home internet access:

• On average, 14% of households have no internet subscription in tracts with no U/L areas.
• But that more than doubles for tracts covered by U/L, increasing to 31% of homes with no internet, on average.

3. CHANGES IN RESPONSE RATES FOR INTERNET FIRST AND INTERNET CHOICE TRACTS

Analysis: has the gap in response rates increased between areas that received Internet First mailings compared with Internet Choice mailings?

When we first analyzed the 2020 Census response rates, we examined tract-level rates as of March 26. On that date, in tracts where households received the Internet First mailing, the average response rate was 29.2%. This was 1 percentage point above the U.S. rate on March 26. But in tracts that received the Internet Choice mailing, the average response rate was 24.6% — 4.5 percentage points below the U.S. rate. Therefore, the gap between response rates for Internet First tracts compared with Internet Choice was less than 5 percentage points.

As of April 16, the average response rate in tracts where households received Internet First mailing was 53.5% — 4 percentage points above the U.S. rate In Internet Choice tracts as of April 16, the average response rate was 44.4%.

Notably, the gap had almost doubled between March 26 and April 16. Not only was the average response rate in Internet Choice tracts 5 points below the U.S. rate on April 16, but the difference between Internet First and Internet Choice tracts is now more than 9 percentage points.

Prepared April 11, 2020

Highlighted topics:

• progress in Update/Leave areas (especially response rates on tribal lands);
• response rates in tracts across cities, by size of city; and
• historically undercounted populations.

Download a presentation of the report below
Download a summary presentation of the report below

---

National Response Rate Trends

The nationwide 2020 Census self-response rate as of Thursday, April 9, was 46.7%. When the Census Bureau first began reporting 2020 response rates on March 20, the U.S. rate was 14.1%. The national rate has continued to increase, but the day-by-day increase in the last week was slower than the first two weeks.

The trend is shown in the chart in Figure 1 below. The chart is annotated to show that from April 8 through April 16, the Census Bureau is mailing paper questionnaires to non-responding households. As part of the Self-Response operation, the Bureau also is sending reminder letters to housing units with mailable addresses that are part of the "Update/Leave" operation. This operation was suspended in March due to COVID-19. Unfortunately, this means that most households covered by Update/Leave have not received their hand-delivered census packets, which contain a paper form and information about how to respond on-line or by phone to the 2020 Census. We will be following the response rates as these paper forms are mailed back to the Bureau and as households in Update/Leave areas respond based on the reminder letters.

The chart visualizes the day-to-day response rates in 2010 (the long red line) compared with the rates so far in 2020 (the shorter red line noted with the arrow). The other lines represent state-by-state response rates in 2020. (The Census Bureau did not publish daily response rates from 2010 at the state level.)

FIGURE 1

NB: Keep in mind the differences in census operations between 2010 and now when comparing response rate trends. Census operations were different in 2010, which may impact the difference in the national trend. For example, there was no online option, there were only two mailings (rather than staggered waves of initial mailings and multiple follow up mailings in 2020), and of course, there was no pandemic in 2010 that caused key census operations to be delayed and completely upended Get Out the Count (GOTC) campaigns that were based significantly on in-person outreach. Nonetheless, one way stakeholders will be measuring the success of the 2020 national self-response operation is whether the nation meets and surpasses its 2010 rate. On a local level, in addition to following the 2020 rate compared with the final 2010 rate, you can also compare how your area is doing compared with others nearby, or similar areas across the state, and for larger regions such as counties, the state, and nation as a whole.

The following graphs show the latest 2020 response rates by state, and the percentage point change by state from March 20, when the Census Bureau first reported response rates, until now.

FIGURE 2 States ranked by April 9 census response rate (the final 2010 U.S. rate and the current U.S. 2020 rate are included for comparison).

FIGURE 3 States ordered left-to-right by lowest response rate to highest on April 9 (including the current national rate and final 2010 response rate for comparison).  Brown bars represent the beginning response rate (March 20); the blue bars represent the increase between 3/20 and 4/9.

Key areas of concern analyzed for Week 3

In addition to the nationwide response rate trends, we focused our Week 3 analysis on the following issues:

1. Progress in Update/Leave areas (especially response rates on tribal lands)
2. Response rates in tracts across cities, by size of city
3. Historically undercounted populations (based on census tract demographic characteristics)
   • Children under age 5
   • People of color
   • Population in poverty
   • Hard-to-count (HTC) tracts
4. Response rates by educational attainment

1. PROGRESS IN UPDATE/LEAVE AREAS

The Census Bureau’s Update/Leave operation encompasses approximately 6.8 million housing units (approximately 5 million of which are stateside; the remaining units are in Puerto Rico). It is focused on communities where using the mail to deliver census invitations might be unreliable because homes lack city-style addresses, or a majority of a community receives mail only at P.O. Boxes, or most residents are seasonal, or a community is recovering from a natural disaster.

In the Update/Leave operation, a census worker visits all housing units, updating the Census Bureau’s Master Address File and verifying the location spatially, and then leaving a census packet at the door. Census workers try to make contact with someone in the home, or nearby, to confirm the address and determine if there are any other living quarters on the property. The Update/Leave packet includes a paper questionnaire, as well as online instructions and a unique ID for online or phone response. Households can submit their census responses online, by phone, or by mailing back the questionnaire.

The Update/Leave operation was suspended on March 18, just days after it began. Only 5% of the housing units covered by this operation had received their census packets (meaning 95% of the housing units covered by Update/Leave have not received their packets yet).

Nationwide, areas with large shares of housing units covered by Update/Leave continue to have low self-response rates. In census tracts across the country that have at least some housing units covered by Update/Leave (only about 20% of tracts nationwide), the average share of homes in Update/Leave is 15%. In tracts above that average, the typical self-response rate is only 25%, well below the U.S. response rate of almost 50%.

Update/Leave on Tribal Lands

For Week 3, we examined how the suspension of Update/Leave is impacting tribal areas, and also whether Update/Leave areas have reliable internet connectivity.

The Census Bureau is publishing response rates for 315 tribal areas across the country. The latest response rates are low for most of these areas: 20% or less in more than half of them. Response rates in some of the largest tribal lands are very low. For example, in the Navajo Nation Reservation and Off-Reservation Trust Land (covering an area of almost 173,000 people spanning three states: Arizona, New Mexico, and Utah), the response rate on April 9 was less than 1% (0.3%). The latest response rate for the Navajo Reservation can be viewed at this link on our map.

The map below in Figure 4 displays the response rates by tribal area in the continental U.S. (There is only 1 tribal area in Alaska for which the Bureau is publishing response rates. Most of the 37 Native Hawaiian areas have response rates above the national average.)

FIGURE 4

Update/Leave and Internet Access

Because most households in the Update/Leave operation have not received their paper forms yet, we wanted to examine the availability of internet access in these areas. We analyzed the share of households that reported having no home internet subscriptions based on the latest American Community Survey estimates (for the 2014-2018 period). We focused on tracts in the 10 states where the Update/Leave share of housing units is greatest (based on our Week 2 analysis: Alaska, Hawaii, Maine, Montana, New Hampshire, New Mexico, South Dakota, Vermont, West Virginia, Wyoming).

The average response rate across these tracts is only 29%, compared to almost 48% response rate for tracts in these states with no Update/Leave housing units. We found that, on average, 21% of households in the Update/Leave tracts in these states have no home internet access.

2. RESPONSE RATES ACROSS TRACTS IN CITIES, BY SIZE OF CITY

Response rates vary substantially between urban areas and non-urban areas; within urban areas, they vary by size of city.  (For this analysis we examined tract-level response rates for tracts that are wholly or mainly within one of the Census Bureau-designated "incorporated places" for which the Bureau is publishing response rates.)

For our analysis, we used a modified scale from the National League of Cities to determine incorporated place or city size by population (all population counts are based on the latest American Community Survey estimates for 2014-2018):

• Small cities have fewer than 50,000 people;
• Medium-size cities have between 50,000 and 300,000 people;
• Large cities have between 300,000 and 1 million people; and
• The largest cities having populations of 1 million or more (there are 10 of these nationwide).

We found that compared with the U.S. rate as of April 9 (46.7%), the average response rate for tracts in the nation's 10 largest cities (with 1 million people or more) was 38%. This is almost 8 percentage points below the national average. In contrast, average tract-level response rates in small (less than 50,000 population) and medium (50,000—300,000 population) cities are modestly above the national average.

For the nation's 10 largest cities with populations of 1 million or more, only two of these cities had response rates on April 9 above the national rate, both of them in California: San Jose and San Diego.  New York, the nation's largest city, had the lowest response rate (36%).

Table 1 below ranks these 10 largest cities by their April 9 response rate, highest to lowest, and provides the latest ACS population estimates as well as each city's final 2010 response rate for context and comparison.

TABLE 1

City	Population estimate	Citywide Response Rate (April 9)	Final 2010 Self-Response Rate
San Jose	1,026,658	54.5	74.6
San Diego	1,401,932	51.6	69.3
Phoenix	1,610,071	44.3	62.9
San Antonio	1,486,521	43.7	67.2
Chicago	2,718,555	40.5	62.4
Dallas	1,318,806	40.2	61.9
Houston	2,295,982	39.3	63.5
Philadelphia	1,575,522	37.9	62.4
Los Angeles	3,959,657	37	68
New York	8,443,713	36	64

Visually, the HTC 2020 map reveals the low response rates across these major cities in comparison to their surrounding suburbs. Figures 5, 6, and 7 below show the tract-level response rate patterns for three of these cities as examples: New York, Philadelphia, and Dallas. Figure 8 shows the difference in tract response rate patterns in San Jose.

FIGURE 5. A concentration of low response rate tracts (in brown) in New York City amid tracts with much higher rates (in blue) across NYC's suburbs. Note, however, that other nearby cities such as Newark to the immediate west and Yonkers just to the north also have concentrations of low response rate tracts. (Click here to view the latest rates online.)

FIGURE 6. A similar pattern between Philadelphia (highlighted in yellow) and its suburbs. You can also see a similar pattern in Trenton, NJ to the northeast. (Click here to view the latest rates online.)

FIGURE 7. Low response tracts concentrated in Dallas and its neighbor, Fort Worth, compared with higher rates outside both cities. (Click here to view the latest rates online.)

FIGURE 8. One of the two exceptions to the trends above: San Jose, CA shown below, with the opposite pattern compared to the three maps above. It has the highest response rate of all 10 cities of 1 million people or more (almost 55% -- higher than the national average). (Click here to view the latest rates online.)

3. HISTORICALLY UNDERCOUNTED POPULATIONS

Children Under 5 Years of Age

Children under 5 represent the age group with the greatest net undercount in the decennial census. Although self-response rates do not tell us the extent of possible undercounting of any specific demographic group, nor do they indicate anything on their own regarding census accuracy, examining response rates in communities where children are at greatest risk of being undercounted can help census stakeholders target their outreach strategies and messaging to promote inclusion of young children in household census responses.

We examined response rates by tracts nationwide across 689 counties that were studied by the Population Reference Bureau, to determine tracts at greatest risk of an undercount of young children.

We found that:

• Tracts with low risk of undercounting young children had an average response rate of 52% (above the national average).
• Tracts with High Risk had an average response rate of 48% (slightly higher than the national rate).
• Tracts with Very High Risk had an average response rate below the national level, at 40%.

We also examined the response rates by risk category for 34 counties that are home to half of all young children in tracts with very high risk of undercount, according to Count All Kids.

Rates are lower for all categories in tracts in these counties:

• Tracts in these 34 counties with low risk of undercounting young kids had an average response rate of 47% (above the national average).
• Tracts with High Risk had response rate of 45% (slightly higher than the national rate)
• Tracts with Very High Risk had an average response rate below the national level, at 39%.

In order to help census stakeholders in these counties, we prepared a table of response rates by county, available at this link. 

People of color

For this analysis, we use ACS estimates for people who reported their race “alone or in combination with other races.” (Note: People of Hispanic origin can be of any race.) We also use ACS estimates for people who identify as being of Hispanic origin, regardless of race.

We examined the latest response rates for tracts based on plurality population for each major racial group and Hispanic origin. We cross-tabulated these tracts based on population size of the city in which they are located, and found the following patterns, discussed below (click to enlarge):

TABLE 2

NOTES: 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.

Also, there are only 222 tracts nationwide in which the plurality population is American Indian/Alaska Native. Most of these (208) are located outside “incorporated places.” There are only 14 tracts within incorporated places, in the smallest city category.

The key findings in Table 2 above are highlighted in red:

• Predominantly Hispanic census tracts had lowest average response rates for cities of all sizes: 35.9%.
• Predominantly Black census tracts also had below-average response rates (for all tracts in those size cities), especially in cities above 1 million people: 34.9%.
• Predominantly Asian census tracts had response rates well above national average in small cities (53.5%) and medium-size cities (52.9%), but well below the national average in cities with more than 1 million people: 38.8%.
• The average response rate for predominantly White census tracts across cities of all sizes was 47.9%.

People with incomes below the poverty line

For this analysis, we focused on what are considered by poverty advocacy groups to be "High poverty" tracts, in which more than 30% of residents for whom the poverty level has been determined have incomes below the poverty line.

The average response rate across all high poverty tracts nationwide is 34.5%, more than 12 points lower than the national average.

Tracts with less than 30% of their population in poverty had an average response rate of 47%, which is above the national average.

"Hard-to-count" tracts

When the CUNY HTC 2020 map was launched in Oct. 2017, it used "mail return rates" from the 2010 Census as the primary metric for tracts at greatest risk of undercount. Unlike real-time response rates that we are tracking now, mail return rates represent the number of households that self-responded, as a share of occupied housing units (households) only. This way of measuring self-response can only be calculated after the census is complete; the calculation removes vacant and nonexistent housing units from the denominator, which can only be identified during the door-knocking phase (Nonresponse Follow-up, or NRFU).

Using the 2010 mail return rate, the HTC map highlights tracts in yellow-to-red on the map as the "hardest to count" tracts in the country, where 73% of households or less mailed back their census form in 2010.

Based on current 2020 self-response rates, these tracts have a lower-than-average response rate of 35% (more than 10 points lower than the U.S. rate).

Tracts that had a higher 2010 mail return rate now have an average self-response rate above the national rate: 49%.

4. RESPONSE RATES BY LEVEL OF EDUCATIONAL ATTAINMENT

We examined response rates by tract compared with education levels for the population age 25 and older. We found:

• Tracts where more people have higher levels of education (college & advanced degrees), response rates tend to be higher (there was a strong positive statistical correlation).
• Conversely, in tracts with greater concentrations of people with a high school degree or less, rates tend to be lower (a strong negative correlation).

Prepared April 5, 2020

Highlighted topics:

• suspension of Update/Leave operation
• possible relationship with COVID-19 trends
• Internet First / Internet Choice / bilingual mailings
• response rates by demographic characteristics

Download a summary presentation of the report below

---

National Response Rate Trends

The nationwide 2020 Census self-response rate as of Thursday, April 2, was 41.3%. When the Census Bureau first began reporting 2020 response rates on March 20, the U.S. rate was 14.1%. It has almost tripled since then.

There was a substantial increase in the day-over-day rate on April 1, i.e., “Census Day” which is the Census Bureau’s “reference day” for filling out census forms (the date is not a deadline – households can self-respond until mid-August). Census stakeholders  across the country organized a concerted digital effort that day to amplify the importance of filling out the census questionnaire. This outreach represented a major shift by the stakeholders. Much of their Get Out the Count (GOTC) strategies and tactics before COVID-19 were based on in-person contact; the public health crisis required stakeholders to pivot quickly to social media, phone outreach, texting, and other “at a distance” methods.

The impact of this one-day outreach campaign is reflected in the chart below. (The chart and figures have been updated with response rates through April 4.) The national self-response rate increased almost 3 points on Census Day over the previous day. Prior daily increases were approximately 2 points per day or less. The chart also shows that the rate increased in most states, reflecting the nationwide outreach effort. We discuss the impact of #CensusDay2020 activities further at the April 3 issue of our HTC Map email newsletter.

FIGURE 1.

The following graphs show the latest 2020 response rates by state, and the percentage point change by state from March 20 till now.

FIGURE 2. States ranked by April 4 census response rate (the final 2010 U.S. rate and the current U.S. 2020 rate are included for comparison).

FIGURE 3. States ordered left-to-right by response rate on April 4.  Brown bars represent the beginning response rate (March 20); the blue bars represent the increase between 3/20 and 4/4. The final U.S. 2010 rate and current U.S. 2020 rate are included for comparison.

Key areas of concern analyzed for Week 2

In addition to the nationwide response rate trends, we have focused our Week 2 analysis on the following issues:

1. Suspension of Update/Leave operation
2. Possible relationship with COVID-19 trends
3. Internet First / Internet Choice / bilingual mailings
4. Response rates by demographic characteristics

1. Suspension of Update/Leave operation

The Census Bureau’s Update/Leave operation encompasses approximately 6.8 million housing units (approximately 5 million of which are stateside; the remaining units are in Puerto Rico). It is focused on communities where using the mail to deliver census invitations might be unreliable because homes lack city-style addresses, or a majority of a community receives mail only at P.O. Boxes, or most residents are seasonal, or a community is recovering from a natural disaster.

In the Update/Leave operation, a census worker visits all housing units, updating the Census Bureau’s Master Address File and verifying the location spatially, and then leaving a census packet at the door. Census workers try to make contact with someone in the home, or nearby, to confirm the address and determine if there are any other living quarters on the property. The Update/Leave packet includes a paper questionnaire, as well as online instructions and a unique ID for online or phone response. Households can submit their census responses online or by phone, or by mailing back the questionnaire.

The Update/Leave operation was suspended on March 20, four days after it began. Only 5% of the housing units covered by this operation had received their census packets (meaning 95% of the housing units covered by Update/Leave have not received their packets yet).

Nationwide, the Update/Leave operation covers a relatively small universe of housing units: the Census Bureau estimates that it only includes 5% of the nation’s residential addresses.  But in some states, the Update/Leave operation covers a substantial portion of homes. For example:

• in Alaska, almost 30% of the state’s housing units were scheduled to receive hand-delivered packets. Alaska’s census self-response rate as of April 2 was 25.5%, the lowest in the country.
• in West Virginia, 27% of the state’s housing units are in Update/Leave areas. As of April 2, West Virginia’s response rate was 29.6%, the second lowest nationwide.
• in New Mexico, almost 18% of the housing units are in Update/Leave areas. The state’s response rate was 30.7% on April 2, the third lowest nationwide.

The shares of housing units presented above are based on the latest housing unit estimates from the 2014-2018 (5-year) American Community Survey data.

Table 1 below presents the 10 states with the largest share of housing units in Update/Leave. The states are ranked in order of lowest to highest self-response rates as of April 2.  All but two of these states have the lowest response rates nationwide.

TABLE 1

The map views below are taken from our HTC map.  The maps compare response rates by census tract as of April 2 with areas where the Census Bureau had planned to hand-deliver census packets to households. Figures 4 and 6 show response rates shaded darkest brown for the lowest rates. Figures 5 and 7 show Update/Leave areas highlighted in yellow. 

Figure 4 shows the response rates from last week in the southwest including Arizona, New Mexico, Oklahoma, and most of Texas.

FIGURE 4

Figure 5 displays the patterns of the Census Bureau’s contact strategies across the same region: households in areas shown in green and purple have received mailed invitations from the Bureau to respond to the census. Households in areas in yellow were to receive hand-delivered packets. Most areas in this region with low response rates overlap very closely with the Update/Leave areas.

Figure 5

In Figures 6 and 7 below, the overlap between low self-response and Update/Leave is more striking. Figure 5 focuses on the Midwest and parts of the Northeast. In particular, tracts in West Virginia, upper Wisconsin, and parts of New York, including the Adirondacks and the Catskills, have especially low self-response rates.

FIGURE 6

Figure 7 shows the same region. The areas in yellow cover almost exactly the same census tracts that are part of the Update/Leave operation.

FIGURE 7

Finally, Figure 8 below uses counties in New Mexico as an example to emphasize the correlation between very low response rates and areas where large shares of households were in the Update/Leave operation. The one outlier county (Los Alamos) has the highest response rate in the state, and a relatively high response rate so far nationwide. Almost all housing units in this country received a mailed invitation from the Census Bureau, and almost all households that responded so far have done so via the internet.

The small circles on the graph represent counties. The red line sloping downward from left to right indicates the negative relationship between response rates and shares of housing in Update/Leave: the greater the concentration of Update/Leave, the lower the response rate.

FIGURE 8

The figures and scatterplot above help confirm that in areas where Update/Leave is prevalent, response rates are, not surprisingly, very low compared to all other tracts. What are some of the relevant population characteristics of these areas?

One population group experiencing historic and persistent census undercounts is people of color. In order to estimate the race and ethnicity (Hispanic origin) characteristics of the population in Update/Leave areas, we examined the tract-level population estimates from the latest American Community Survey (for the 2014-2018 period) for tracts in the 10 states listed in Table 1 above (with the greatest shares of housing units covered by Update/Leave).

We calculated that on average, 17% of a tract’s housing units in these states was covered by Update/Leave. We then tallied the population estimates by race/ethnicity for tracts that were above average (i.e., where more than 17% of housing units were in Update/Leave). This represents an overall estimated population of approximately 2.76 million people (the total population of these “above average” tracts, not only people living in Update/Leave housing units).

Of these 2.76 million people:

• 84% are White (people identifying as White on the ACS questionnaire either alone or in combination with other races, and without regard to Hispanic origin);
• 11% are American Indian (also “alone or in combination” with other races regardless of Hispanic origin);
• 9% are Hispanic (people of Hispanic origin can be of any race); and
• 3% are Asian (“alone or in combination” with other races regardless of Hispanic origin); and
• Other racial groups comprise smaller percentages.

These statistics portray the racial and ethnic composition of the population overall in the above-average Update/Leave communities in those 10 states. Nationwide, however, the Update/Leave operation is primarily concentrated in American Indian communities. Tracts within American Indian reservations fall under the Update/Leave operation, unless the tribe elected to be counted through the separate Update/Enumerate operation, which doesn't involve self-response. Therefore, suspension of Update/Leave means that most households on American Indian reservations have not received their census packets yet.

2. Possible correlation between Census self-response and COVID-19 incidence

We downloaded county-level data from the New York Times for the number of COVID-19 cases and COVID-19 related deaths in the U.S., as of March 31.  We also separated cases in New York City by county within the city.

Comparing these data with county-level Census self-response rates shows a very slight positive correlation between self-response rates and COVID-19 cases, as well as response rates and COVID-19 related deaths.

It may be that because COVID-19 incidence is distributed across the country, there may not be enough geographic differentiation in its impact to show an easily identified correlation with census response. We will continue to examine this relationship.

3. Census response rates in areas receiving Internet First and Internet Choice mailings, and areas receiving bilingual mailings

The Census Bureau has sent mailings to most households across the country to invite participation in the 2020 Census.  Most households nationwide (80% of the mail-out universe) received an “Internet First” mailing, which included a letter of invitation to respond online with a unique household ID number, as well as a Language Assistance sheet with 13 phone numbers to ask questions and to respond by phone.

Other households (20% of the mail-out universe), in areas lacking reliable internet access or where there are large concentrations of householders age 65 or older, received “Internet Choice” mailings. These census packets included the same materials as Internet First packets, but also included a paper questionnaire, giving households the “choice” to respond by mail from the start.

As of April 2, in tracts where households received an Internet First mailing:

• The average overall response rate was 42.9%
• The average internet response rate was 42.1%
  • 98% of the response in “Internet First” tracts has been online.
  • Both rates are above the national response rate.

In Internet Choice tracts, as of April 2:

• The average overall response rate has been lower than Internet First tracts and nationwide rate: 35.4%.
• The average internet response rate was 16.7%, In other words, less than half of the response in Internet Choice tracts has been online.

For both types of mailings, the Census Bureau further identified census tracts where at least 20% of households were recognized as needing "Spanish assistance.” These households received either an Internet First or Internet Choice mailing with bilingual English/Spanish materials.

Our analysis of response rates in “bilingual 2020 Census tracts” as of April 2 shows the following:

In Internet First tracts with bilingual mailings:

• The average overall response rate was 33.2%.
• The average internet response rate was 32%.

In Internet Choice tracts with bilingual mailings:

• The average overall response rate was 30.6%.
• The average internet response rate was 16.8%.

4. Response rates analyzed by tract-level population characteristics

Self-response rates do not reveal any information about the number of people or the share of population that has been counted; self-response only represents households that have responded, as a share of all housing units on the Census Bureau’s address list. Self-response rates also do not tell us anything about the characteristics of the population or the households that have been counted so far, such as their race or ethnicity or whether there are young children in the household.

But we can examine patterns of response rates based on tract-level population characteristics, using the latest American Community Survey (ACS) estimates, to gauge how well communities with historically undercounted populations are responding compared with other communities. Below are selected findings from our Week 2 analysis.

Race/ethnicity

For our Hard to Count map, we use ACS estimates for people who reported their race “alone or in combination with other races.” (Note: People of Hispanic origin can be of any race.) We also use ACS estimates for people who identify as being of Hispanic origin, regardless of race.

Using these estimates, we compare tract-level self-response rates for tracts whose populations are predominantly one race category or another, “alone or in combination with other races.” Note that we did not include tracts that are predominantly “American Indian or Alaska Native” or “Native Hawaiian or Pacific Islander.” Those tracts largely fall within the Update/Leave operation, which has been suspended.

Nationwide at the census tract level, as of April 2:

• Tracts whose population is predominantly Hispanic had the lowest response rate, on average, compared with the other population groups we examined: 30.5%.
• The average response rate for tracts whose population is predominantly Black is 35%.
• The average response rate for tracts whose population is predominantly Asian is 41.8% (above the national rate).
• Tracts whose population is predominantly White had the highest average response rate: 42.5% (above the national rate).

For other population groups, we were able to compare general response rate trends, including:

• Are the rates generally higher or lower if a tract’s population has a smaller or greater share of one demographic group or another; and
• How strong is that correlation: strongly positive or negative, or moderately or “weakly” positive or negative?

The results in Week 2 are similar to Week 1. Self-response rates tend to be lower as the share of foreign-born persons, people with incomes below 200% of the poverty level, limited English proficiency households, and renter households, viewed separately, increase at the tract level.

For Week 2, we also began to examine the trend in response rates compared with tract-level population of children under age 5, who had the highest net undercount of any age group in the 2010 Census.

For tracts that have a greater than average share of children under 5 years of age, the average response rate as of April 2 is 40.2% (slightly less than the national average).

For Week 3, we will examine response rates in tracts that are considered to have a High or Very High risk of undercounting for children under 5, based on an analysis by the nonprofit, nonpartisan Population Reference Bureau.

Prepared March 30, 2020

Highlighted topics:

• internet response patterns & Internet First and Internet Choice
• bilingual mailings
• historically undercounted populations

Download a summary presentation of the report below

---

National Response Rate Trends

The nationwide 2020 Census self-response rate as of Thursday, March 26, was 28.1%.

This represented an increase of 14 percentage points since the first day the Census Bureau began publishing response rates on March 20, when the national rate was 14.1%. Overall, this increase represents an average daily increase of 2 percentage points. This rate of increase was slower than over a comparable time period in 2010. In 2010, during the first week that response rates were published, the nationwide rate increased from 12.5% on the first day to 39.3% at the end of the first week – an average increase of 3.8 percentage points per day.

The chart below visualizes the increase in 2010 (the long red line) compared with the increase so far in 2020 (the shorter red line noted with the arrow). The other lines represent state-by-state response rates in 2020. (The Census Bureau did not publish daily response rates from 2010 at the state level.)

The following tables list the 2020 response rates so far by state, and the percentage point change by state from March 20 through March 26.

TABLE 1. States ranked by March 26 census response rate (nationwide rate included for comparison)

State	Response Rate (%) as of March 26
Wisconsin	35
Minnesota	34.5
Iowa	34.3
Nebraska	34.2
Michigan	32.8
Kansas	32.2
Virginia	31.7
Idaho	31.6
Utah	31.5
Illinois	31.1
Indiana	31
Ohio	30.5
Washington	30.2
Massachusetts	29.9
Missouri	29.8
South Dakota	29.8
Mississippi	29.7
Pennsylvania	29.6
Alabama	29.5
Oregon	29.4
Connecticut	29
Maryland	29
Tennessee	28.9
Colorado	28.9
Kentucky	28.3
North Dakota	28.2
Nationwide	28.1
Nevada	28.1
Florida	27.9
Arkansas	27.8
Rhode Island	27.8
Louisiana	27.5
New Jersey	27.5
Arizona	27.5
Delaware	27.4
California	27.3
New Hampshire	26.8
Georgia	26.1
North Carolina	25.8
District of Columbia	25.7
South Carolina	25.6
Oklahoma	25.2
Texas	23.9
New York	23.9
Hawaii	23.8
Maine	22.8
Vermont	22.6
Montana	22.1
Wyoming	21.4
New Mexico	20.4
West Virginia	19.1
Alaska	16.2

TABLE 2. States ranked by percentage point change in census response rate, March 20 through March 26 (nationwide increase is included for comparison)

State	Percentage Point Increase March 20 to March 26
Minnesota	18.1
Michigan	17.3
Wisconsin	17.2
Utah	16.9
Washington	16.6
Ohio	15.9
Connecticut	15.8
Maryland	15.4
Iowa	15.4
Virginia	15.3
Indiana	15.3
New Hampshire	15.3
Oregon	15.1
Tennessee	15.1
Nebraska	14.9
Illinois	14.8
Kentucky	14.7
Georgia	14.7
Massachusetts	14.7
Mississippi	14.7
Louisiana	14.7
Alabama	14.4
Colorado	14.4
Pennsylvania	14.3
North Carolina	14.2
North Dakota	14.2
New Jersey	14.2
Missouri	14
Nationwide	14
Kansas	13.9
Idaho	13.8
Florida	13.6
South Carolina	13.4
Arkansas	12.9
Delaware	12.8
Texas	12.7
Nevada	12.6
California	12.6
Arizona	12.6
Hawaii	12.2
South Dakota	12.2
New York	11.7
Montana	11.5
Oklahoma	11.5
Rhode Island	11
District of Columbia	10.9
Maine	10.7
Wyoming	9.9
Vermont	9.7
New Mexico	9.7
West Virginia	9.3
Alaska	8.6

The rate of increase in response rates in the first week of reporting was uneven across the country. The following map shows the county-level changes in response rates from March 20 through March 26, in percentage points.

The red ovals on the map highlight three areas as examples of especially modest increases: West Virginia, the Texas border, and the Arizona/New Mexico border region. But the next map shows that these areas also are where the Census Bureau was planning to hand-deliver census packets to households as part of its “Update/Leave” operation (the areas highlighted in yellow on the map below). This operation was suspended shortly after it started on March 16 due to COVID-19, meaning most households in those areas have not yet received their invitation to respond to the census on-line, by phone, or using the paper form included in those packets.

While households in the Update/Leave universe can respond to the census on-line or by phone with the unique ID number assigned to their address, many households in those areas might lack reliable broadband access and even phone service, and they might not be aware of this “non-ID response” option. Among the communities affected by the suspension of this operation, which covers about 6.6 million housing units: most American Indian reservations, some colonias in the Rio Grande Valley, all of Puerto Rico, and other communities lacking city-style addressing or direct mail delivery.

Internet-only Response Rate Trends

A major innovation of the 2020 Census is the option to fill out the census questionnaire online, in English plus 12 non-English languages. The Census Bureau is hoping most households use this online option in order to save time and costs. (Putting aside households in the Update/Leave operation and a few other “remote” areas, only 20 percent of all other households nationwide received a paper form in the first mailing. All households also can give their census responses by phone, in the same 13 languages as on-line, plus TDD for hearing impaired persons.)

The online response rate nationwide as of March 26 was 24.5%. That means 87% of self-response nationwide has been via the Census Bureau’s online portal.

The online response rate pattern was uneven across the country. The following map shows the online response patterns by census tract nationwide, as a share of overall response rates. In tracts shaded in dark blue, 80% or more of the responses have been via the online portal. In the lighter green census tracts, fewer than 40% of the responses have been online, meaning most of the responses have been either by phone or by mail. (The Census Bureau does not report a separate rate for phone or mail responses.)

The red ovals on the map highlight four regions where internet response rates are relatively low — in other words, where most responses are by phone or by mail. A likely explanation for this pattern is that households in these areas received the “Internet Choice” packets, which contained the paper questionnaire, as well as information on how to fill out the census form online or by phone. The map below shows census tracts, highlighted in light green, where households received Internet Choice packets.

These geographic correlation between type of census mailing and online response rates are reflected also in the national statistics. As of March 26, in tracts nationwide where households received “Internet First” mailings, which did not include a paper questionnaire:

• average overall response rate was 29.2%; and
• average internet response rate was 28.6%.

In other words, almost 98% of self-response in “Internet First” tracts has been online.

As of March 26, in tracts nationwide where households received “Internet Choice” mailings:

• average overall response rate was 24.6%; and
• average internet response rate was 10.6%.

Two findings are noteworthy about the response rates in “Internet Choice” tracts. First, on average, most households in census tracts that received the paper questionnaire as a “choice” along with internet and phone options chose not to fill out the census form online and instead either mailed back the paper form or called in their responses by phone.

Second, the average overall response rate in these “Internet Choice” tracts was lower than in “Internet First” tracts.

Bilingual Mailings

For the 2020 Census, the Census Bureau also identified census tracts where at least 20% of households speak Spanish as their primary language at home, based on estimates from the ongoing American Community Survey. Households in these tracts received “bilingual English-Spanish” materials, either in their Internet First or Internet Choice census packets. Our analysis of response rates in “bilingual 2020 Census tracts” over the first week of reported response rates (as of March 26) showed the following:

In “Internet First” tracts that received bilingual mailings:

• average overall response rate was 21%
• average internet response rate was 20.5%

In “Internet Choice” tracts that received bilingual mailings:

• average overall response rate was 19.4%
• average internet response rate was 9.9%

Response rates analyzed by tract-level population characteristics

It’s important to remember that self-response rates do not reveal any information about the number of people or the share of population that has been counted; self-response only represents households that have responded. Self-response rates also do not tell us anything about the characteristics of the population or the households that have been counted so far, such as their race or ethnicity or whether there are young children in the household.

But we can examine patterns of response rates based on tract-level population characteristics, using the latest American Community Survey (ACS) estimates, to gauge how well communities with historically undercounted populations are responding compared with other communities. Below are selected findings from our Week 1 analysis.

Race/ethnicity

For our “Hard to Count” map, we use ACS estimates for people who reported their race “alone or in combination with other races.” (Note: People of Hispanic origin can be of any race.) We also use ACS estimates for people who identify as being of Hispanic origin, regardless of race.

Using these estimates, we compare tract-level self-response rates for tracts whose populations are predominantly one race category or another, “alone or in combination with other races.”

Nationwide at the census tract level, as of March 26:

• The average response rate for predominantly Hispanic tracts is 18.8%, the lowest of all groups that we examined.
• The average response rate for tracts whose population is predominantly Black is 23.4%.
• The average response rate for tracts whose population is predominantly Asian is 28.1%.
• Tracts whose population is predominantly White had the highest average response rate: 29.1%.

For other population groups, we were able to compare general response rate trends, including:

• Are the rates generally higher or lower if a tract’s population has a smaller or greater share of one demographic group or another; and
• How strong is that correlation: strongly positive or negative, or moderately or “weakly” positive or negative?

This week, we prepared this analysis for the following characteristics:

• Foreign born population
• Households with limited English proficiency
• People with incomes below 200% of the poverty line
• Renter-occupied households

Foreign born population

Nationwide at the census tract level:

• Response rates tend to be lower in tracts with a greater share of population that is foreign born (this is a modest relationship; the correlation statistic = -.218).
• In census tracts in large cities in metro areas, the pattern is stronger than the overall nationwide trend:
  • the correlation statistic is -.353, indicating that it’s more likely in metro cities that response rates are lower in communities with a greater share of immigrants.

LEP / low income / renters

Limited English proficiency (LEP) households:

• Response rates tend to be lower in tracts with a greater share of LEP households (correlation statistic = -.325)
• In census tracts in large cities in metro areas, the pattern is stronger:the correlation statistic is -.428

Low income population:

• Response rates tend to be lower in tracts with a greater share of people w/low incomes (incomes below 200% of poverty level); the correlation statistic = -.526.
• In metro cities, the pattern is also strong: correlation statistic = -.582

Renter households:

• Response rates tend to be lower in tracts with a greater share of renters; the correlation statistic = -.415
• In metro cities, the pattern is much stronger: the correlation statistic = -.610