NPS8_Supporting Statement_Part B-2020

B. Collection of Information Employing Statistical Methods

1. Universe and Respondent Descriptions

The NPS-8 data collection is a complete enumeration of persons under sentence of death and a review of the status of capital statutes in the United States. BJS collects data regarding inmates under sentence of death from state departments of corrections (DOCs) in jurisdictions that have capital sentencing laws in effect and the Federal Bureau of Prisons (BOP). Because state and federal law determines eligibility for a sentence of death, BJS also collects information on the status laws governing how jurisdictions apply and carry out of death sentences from the attorney general in each state, the U.S. Attorney for the District of Columbia, and the Office of General Counsel at BOP. Tracking the changes to capital statutes provides contextual information essential to understanding changes in the population of prisoners under sentence of death.

For the NPS-8, BJS includes all state DOCs having legal authority to hold prisoners under sentence and the BOP to establish the universe of respondents for obtaining data on persons under sentence of death. As of year-end 2018, this included 34 DOC respondents and one from BOP, who had responsibility for 2,628 prisoners reported under sentence of death. Conducting a full enumeration each year allows for the collection of complete information. Sampling to obtain these data would impede BJS’s ability to achieve its goals for this collection and would greatly diminish the quality and utility of the data. One important objective of the collection is to provide jurisdiction-specific data on prisoners under sentence of death and to track the movement of individuals entering and exiting death row by jurisdiction.

Sampling from the roster of persons under sentence of death to obtain data to generate precise estimates of the characteristics of all persons under sentence of death by jurisdiction is not feasible given the distribution of persons under sentence of death by jurisdiction. At year-end 2018, for example, 26 states and the BOP held fewer than 90 persons under sentence of death. The distribution of important attributes of prisoners under sentence of death (e.g., age, race, Hispanic origin, sex) would require complex stratification designs, with some cells still requiring complete enumerations (e.g., for sex). Even in jurisdictions with relatively large numbers of persons under sentence of death (e.g., in 2018, California with 736 or Florida with 344), sampling would result in imprecise estimates of movements of inmates removed from under sentence of death by means other than execution.

The universe of NPS respondents is as follows:

a. NPS-8/NPS-8A respondents

Respondents are comprised of staff from the DOC in each state that authorizes capital punishment and from BOP. The appropriate respondents are designated by the individual agency. While some respondents are in the records office of the DOC, others are in the facilities responsible for the custody of the inmates under sentence of death. For the 2019 reference year, there were 35 NPS-8/NPS-8A respondents in the 35 jurisdictions which authorized capital punishment.

b. NPS8-B/NPS-8C respondents

Respondents are comprised of legal staff from the Office of the Attorney General in each state, the U.S. Attorney’s Office for the District of Columbia, and the Office of General Counsel at BOP. Each respondent will be asked to provide information about changes to the capital statutes during the reference year. Based on the status of death penalty statutes in the jurisdiction on December 31 of the previous year, a respondent will be asked to complete the NPS-8B (Status of Death Penalty - No Statute in Force) or the NPS-8C (Status of Death - Statute in Force). For the 2019 reference year, 17 respondents completed the NPS-8B and 35 respondents completed the NPS-8C.

Universe and response rate
	Universe		Expected response rate
Departments of Correction and BOP	35	100%
Attorneys General, BOP, and the U.S. Attorney for the District of Columbia	52	100%

For the 2019 reference year, as in previous years, no respondents refused to participate. Based on response rates for previous reference years, it is expected that 100% of respondents from the offices of the attorneys general, the U.S. Attorney for the District of Columbia, BOP, and the state DOCs will continue to participate in the NPS-8 collection.

2. Procedures for Collecting Information

In late November, Census Bureau staff will contact each respondent from the prior year to verify their contact information in November of the reference year.

a. Departments of Correction and BOP

At the end of December of the reference year, materials will be mailed out to all respondents (for sample introductory letter, see Appendix 7). For NPS-8/8A respondents who request the mail-out/mail-back option, NPS-8 forms will be sent, as needed, in order for respondents to report inmates received under sentence of death from January 1 to December 31 for the current reference year (see Appendix 3). Respondents who reported inmates under sentence of death as of December 31 of the previous collection year will be sent an NPS-8A form (see Appendix 4) pre-filled with the previously submitted data for each inmate. Respondents indicating a preference for the web-based data collection tool will be mailed one NPS-8 form and a blank NPS-8A form to provide a reminder of the information BJS has requested in the previous collection years.

In a separate mailing, all NPS-8/8A respondents will be sent a username and password to access the web-based data collection tool and a flyer with instructions for submitting their forms (Appendices 8 and 9). This flyer will also be sent to respondents who requested the mail-out/mail-back option in the event that they decide to try the web-based data collection tool.

Respondents will be given a submission date based on the size of the population of inmates under sentence of death in their jurisdiction, with the latest requested submission in mid-March. Typically, those with smaller population will be asked to provide their responses by mid-February, and those with a moderate population of death row inmates will be given an extra week (around February 22). Respondents in jurisdictions with a large population of inmates under sentence of death will have their submission date extended through mid-March to allow for sufficient time to review, update, and submit records.

b. Attorneys General, BOP, and the U.S. Attorney for the District of Columbia

NPS-8B or NPS-8C forms (see Appendices 5 and 6) will be sent to staff in the Attorney General’s office of each state, BOP, and the U.S. Attorney for the District of Columbia. Forms will be sent either by regular mail or by e-mail, based on the stated preference of each respondent. Each respondent will receive an introductory letter, which includes a username and password for them to access their forms via the web-based tool (Appendix 10), a flyer with instructions for submitting their forms (Appendix 11), a form, and a return envelope. The NPS-8B/8C respondents will be asked to submit their responses by mid-February.

If respondents submit paper forms, (generally mid-January to late April), data will be entered in the database by Census Bureau staff and reviewed. For respondents who submit data electronically, Census Bureau staff will receive a notification that records have been updated, and these updates will be downloaded from the website and incorporated into the database. Upon reviewing the changes, follow-up phone calls will be made to the respondents by Census Bureau staff, as necessary, and the data will be corrected as appropriate.

In the summer following the report year, the Census Bureau will deliver two data files to BJS: one file will contain information about persons under sentence of death at any time during the report year (active file); the other will provide information about persons who were previously under sentence of death (historical file).

Prior to the release of the statistical tables, BJS sends a letter to respondents to notify them of the expected release date and includes a courtesy copy of all state-specific tables for their review (Appendices 12 and 13).

3. Methods to Maximize Response

As BJS has consistently obtained a 100% unit response rate for this collection, the methods used in prior years’ collections will continue to be followed, while at the same time seeking ways to reduce burden.

As the Census Bureau, BJS’s data collection agent for the NPS-8, begins to receive the submissions for a given reference year (mid-January to late April), follow-up calls will be made to clarify any inconsistencies in the information submitted. As necessary, follow-up phone calls will be made to respondents who have not returned forms. As part of the follow-up activities, the Census Bureau will respond to queries from respondents about the data collection and instruments, as well as any inquires about difficulties they may have encountered during the data collection cycle.

4. Testing Procedures

BJS has consistently achieved 100% unit response rate for this data collection. In addition, for critical items such as race, sex, and date of capital sentence, BJS has obtained 100% item response rates. The highest item non-response rates appear in variables that are unrelated to correctional administration, but, even among these, item non-response rates are comparatively low. For example, for the 2018 collection, marital status at the time of imprisonment was missing for 14% of cases and legal status at the time of the capital offense was missing for 11% of cases. The highest item non-response was for education level, with 21% of the cases having missing data.

BJS is not planning to implement modifications to the collection during this upcoming three-year cycle. Hence, there are no new items to be tested. Beginning with the 2010 reference year, BJS moved the NPS-8 collection from the Demographic Surveys Division to the Criminal Justice Branch of the Economic Reimbursable Surveys Division of the Census Bureau. This transition prompted BJS and the Census Bureau to undertake a complete redesign of the web-based collection system. The restructured system is more user-friendly than previous online data collection efforts. The current system includes several features suggested by respondents that make it easier for them to navigate and report their data.

The Census Bureau regularly engages respondents during the data collection cycle to inquire about problems or suggested revisions. Both users and non-users of the web-based data collection tool will continue to be asked about the reasons for the data submission mode they elect to use. In addition, respondents will be asked to describe their experiences using the web-based tool and to offer any suggestions for changes/ improvements.

Since the initial implementation of a web-based data collection in 2009, BJS has observed a dramatic increase in the number of responses submitted electronically:

NPS-8/8A electronic submissions increased from 5% of all responses in 2008 to 100% in 2019; and NPS-8B/8C electronic submissions increased from 6% in 2009 (the first year it was available to these respondents) to 100% in 2018.

5. Contacts for Statistical Aspects of the Data Collection

The Corrections Statistics Unit at BJS takes responsibility for the overall design and management of the survey, including the development of the questionnaires, and the analysis and publication of the data. BJS contacts include:

Tracy L. Snell, Statistician

Corrections Unit

Bureau of Justice Statistics

810 7^th Street, NW

Washington, D.C. 20531

(202) 616-3288

Richard Kluckow, Chief

Corrections Statistics Unit

Bureau of Justice Statistics

810 7^th Street, NW

Washington, D.C. 20531

(202) 598-0597