This memorandum documents updates made to the Nonresponse Followup (NRFU) operation since the release of 2020 Census Detailed Operational Plan for: 18. NRFU Operation on July 15, 2019.

Many changes were made to the NRFU operation in response to the COVID-19 pandemic, including changes to training content, the operation schedule, and the scope of NRFU operational components. This document identifies these changes in two parts: Major Changes and Minor Adjustments. Nearly all of these changes are the result of the unprecedented circumstances impacting the 2020 Census, including shutting down operations for the pandemic and natural disasters such as hurricanes and wildfires. While there were many notable changes to the operation, the originally published detailed operational plan (DOP) still largely represents the core components and best practices that were implemented.

Descoped Components

The Early NRFU component was scheduled to begin in April 2020, which coincided with the initial phase of the COVID-19 pandemic. To protect the health of the public and Census Bureau staff, the Early NRFU component was descoped. This allowed time for the Census Bureau to implement several safety measures prior to starting field operations for the NRFU operation. Sections and tables in the NRFU DOP that reference Early NRFU are no longer valid. Those instances are outlined below:

• Section 2.2.2.3, page 8.
• Section 2.3.1, pages 19-20.
• Section 2.3.4.3, pages 36-37.
• Section 2.3.4.4, page 37.
• Table 7, Section 2.7.3, page 64.
• Section 3.1.3, pages 88-89: Identification of Early NRFU areas did happen as part of the operational design. However, the Early NRFU data collection from April 9 through mid-May mentioned in this paragraph did not occur during those dates or any other time.
• Section 3.2.2.2, page 96: The Early NRFU universe was combined with the NRFU production universe. Enumeration of Early NRFU housing units occurred on the same timeline as all other NRFU production cases.
• Table 13, page 126.

As noted in the NRFU DOP, Manager Visits (MV) cases were intended to group cases in large multiunits to allow enumerators to work with property managers to determine the housing unit statuses. However, changes to the operational timeline and workload creation were necessary in response to the COVID-19 pandemic. Because of the changed workload creation timeline, multiunit cases were not grouped correctly. The resulting manager visit cases omitted some cases. In particular, cases that had administrative records were placed in a separate workload and not included in multiunit groupings. This negated the efficiency of conducting manager visits because they were not included in the list of units to the manager. As a result, the Census Bureau decided to descope MV cases prior to the start of NRFU production. The manager visit cases were closed by the Field Operational Control System, which then released the individual housing units in the multiunit buildings for data collection. Sections and tables in the NRFU DOP that reference Manager Visits and multiunits are no longer valid for 2020, but manager visits remain a priority for the 2030 NRFU operation. Those instances are outlined below:

• Section 2.2.2.3, page 6.
• Section 2.2.2.3, page 7.
• Table 1, page 12.
• Section 2.2.4.1, page 13.
• Section 2.2.4.1, page 14.
• Table 2, page 16.
• Section 2.2.4.2, page 18.
• Section 2.3.1, page 20.
• Section 2.3.4, page 33.
• Section 2.3.4.2, pages 34-36.
• Section 2.6, page 55.
• Section 3.2.2.3, page 97.
• Section 3.3.1, page 98.
• Section 3.3.1, page 99.
• Section 3.3.3, pages 100-103.
• Appendix C, page 132.

Operational Changes

In response to the COVID-19 pandemic, the NRFU operation made a number of changes to enumerator training content, logistics, and timing. A module was added to the training that explained what COVID-19 was and ways to minimize its spread. Training was also shifted to correspond with the updated operational timeline (see updates below). To minimize in-person contact, the orientation classroom training day was reduced from a full day to a 2-hour appointment. The remainder of the orientation day and the capstone training day were conducted through self-study reading, a podcast, and conference calls. In between the orientation and capstone days, staff took online training at home, as originally planned. The DOP sections impacted by these changes are:

• Table 1, page 12: As noted above, classroom training did not occur, and telephone sessions and podcasts were added to the scope of training.
• Section 2.2.4, page 19.
• Section 2.2.5, page 19: Interactive scenarios for learning how to handle difficult and unusual situations that are a major component of the capstone did not happen in-class, instead it happened as part of the telephone capstone day.
• Section 2.5.1, page 49.
• Section 2.5.3, page 49.
• Section 3.1.6.1.1, page 91.

In the early stages of the NRFU operation, it was clear that some areas could not be contacted in person for many reasons, including state and local government restrictions for in-person contact because of the COVID-19 pandemic, expansive wildfires on the west coast, and an active hurricane season in gulf states. In an effort to gather as much data as possible, the NRFU operation prepared and implemented telephone follow-up as part of the contact strategy. Details about the telephone strategy are outlined below:

• Phone numbers were extracted from the Census Bureau’s Contact Frame, which links phone numbers to addresses in the Master Address File. Any phone number in the Contact Frame connected to an address of a case in the NRFU universe was attached to the NRFU case.
• Enumerators in areas with high COVID-19 rates, impacted by natural disasters, or with low completion rates were allowed to use telephone follow-up to resolve cases. Cases attempted by telephone were considered contact attempts for purposes of calculating proxy eligibility and use within the administrative records modeling criteria.

The telephone strategy was an unplanned contingency in 2020. Analysis will be conducted on its effectiveness and whether a telephone strategy should be a planned contingency in the future. The DOP section impacted by this change is Section 2.2.2.3, page 6: The telephone strategy outlined above should be included in the list of basic data collection procedures.

In reference to the data that feed administrative record modelling of NRFU cases, the date of availability of tax data from the Internal Revenue Service (IRS) changed because of changes in the IRS filing deadline (from April 15, 2020, to July 15, 2020) and the NRFU start date, both of which were because of the COVID-19 pandemic. This affected the following section of the NRFU DOP:

• Section 2.3.2, page 24: The IRS data were updated using multiple rounds on July 15, August 10, and September 3.

Inputs to the NRFU supplemental workload/file were expanded after the completion of the In-Field Address Canvassing operation. The Census Bureau decided to include in this universe cases that were identified as a “Delete” during the Address Canvassing operation, as well as housing units that were discovered during initial contact for the group quarters and transitory locations living quarters enumerations. That update impacts the following section of the NRFU DOP:

• Section 2.3.3.6, page 29.

In the original design, Phase 2 and the Closeout Phase of the NRFU contact strategy included case completion and date criteria that governed the start of the phases. The initial design included a “no earlier than” date that governed when census field supervisor (CFS) areas could enter the phase. The design also included a date when all CFS areas became eligible to begin the next phase.

During the operation, many CFS areas became eligible for these phases sooner than expected. The Census Bureau set the eligibility criteria based on case completion results from the 2018 End-to-End Census Test, but during 2020 operations, CFS areas were meeting the completion criteria sooner than they had during the test. In some instances, CFS areas would become eligible and still have many cases that had not received a first attempt. To account for this, the Census Bureau updated the completion thresholds for both phases in early August. The Phase 2 criterion was updated from 60 percent to 85 percent, and the Closeout Phase criterion was updated from 85 percent to 90 percent. These changes affect the following sections of the NRFU DOP:

• Section 2.3.1, page 21.
• Section 2.3.1, page 22.

In reference to the date criteria, the soft launch of NRFU meant that ACOs began operations at varying dates in the operational life cycle. As a result, the “no earlier than” date was too far out for ACOs that started data collection earlier in the cycle. Since the “no earlier than” date could not be modified for each ACO, the Census Bureau decided to remove that criteria completely at the beginning of the NRFU operation. Given the shift in operational timeline, the Census Bureau also replanned the dates when all CFS areas would become eligible for Phase 2 and the Closeout Phase. The new date of eligibility was September 4, 2020, for Phase 2 and September 11, 2020, for the Closeout Phase. These changes impact the following sections of the NRFU DOP:

• Section 2.3.1, page 21.
  
• Section 2.3.1, page 22.

When assigning cases, the NRFU operation used Best Time to Contact probabilities derived from administrative records. However, during the operation, using that parameter slowed the assignment of cases to field staff. The contact probabilities were eliminated on August 12 to allow for more cases to be automatically assigned. That change impacts the following sections of the NRFU DOP:

• Section 2.2.4.2, page 18.
• Section 2.3.2, page 24.

In early September, the Census Bureau was anticipating direction from Congress to complete the 2020 Census apportionment count by the legal deadline of December 31. In an effort to meet that legally mandated deadline, the operation implemented a series of changes aimed at reducing the timeline of NRFU. Shortly after the implementation of those changes, a federal judge issued a court order that reversed those changes. One of those changes was to the number of contact attempts made for NRFU Reinterview (NRFU RI) and Self-Response Quality Assurance (SRQA) cases. Attempt days were first reduced from six to three. This situation existed for approximately two weeks before the court order resulted in the criteria being set back to six. During that time period, some NRFU RI and SRQA cases were closed after three attempts and processed as such. The majority of NRFU RI and SRQA cases did receive six attempts, however, this nuance should be noted when considering the following section in the NRFU DOP:

• Section 2.2.2.3, page 7.

The creation of Phase 2 and the Closeout Phase of the NRFU contact strategy was largely predicated on mid-decade research findings that indicated that high-performing enumerators were a significant factor in completing difficult cases. However, during the NRFU operation, federal court orders blocked the Census Bureau from reducing field staff, including those who were low-performing. That limitation impacts the following section of the NRFU DOP:

• Section 2.3.6, page 43.

The onset of NRFU data collection was rolled out in stages by groups of area census offices (ACOs) based on COVID-19 restrictions/virus spread and a variety of administrative factors. A soft launch of six ACOs, one in each region, began on July 16 (Cycle 1A). Another six ACOs (Cycle 1B) began data collection on July 23. Cycle 2, with 32 stateside ACOs and the three ACOs in Puerto Rico, began on July 30. Cycle 3 had 39 ACOs and began on August 3. The remaining ACOs started data collection on August 9.

A revised Table 3 of NRFU dates is presented below with the planned and actual dates. Additional references to dates that became inaccurate in the DOP are:

• Section 2.2, page 2.
• Section 2.2.1, page 4.
• Section 2.2.2.3, page 8.
• Section 2.3.2, page 23.

Table 1: NRFU Key Activities and Dates