SPD Working Paper SPD99-3 SPD99: Proposal to provide incentives in the upcoming 1999 Survey of Program Dynamics

MEMORANDUM FOR Dr. Nancy Kirkendall
Office of Management and Budget

From: Alan R. Tupek
Chief, Demographic Statistical Methods Division

Subject: SPD99: Proposal to Provide Incentives in the Upcoming
1999 Survey of Program Dynamics - Revised

I. Executive Summary

In this document, we suggest an alternative proposal concerning incentive use in the upcoming data collection of the 1999 Survey of Program Dynamics (SPD). The original proposal referred to in this document was an attachment to a letter addressed to you from Chester E. Bowie, Chief of Demographic Surveys Division, dated February 5, 1999. The revised proposal reflects our understanding of the outcome of the discussion we had with you and Katherine Wallman on the use of incentives in the 1999 SPD. An overview of the changes from the original proposal is given in the following table. The detailed revised proposal is given below.

Overview of 1999 SPD Incentive Proposal Changes

Change # Original Proposal Alternative Proposal

1 every SPD household eligible would receive an incentive every Type A noninterview household from >98 SPD data collection and any new potential Type A in the >99 collection would receive an incentive

2 incentive is a $20 debit card, delivered in the advance letter by priority mail incentive is a $40 debit card, delivered by priority mail in either the advance letter for >98 Type A or the refusal letter for potential >99 Type A.

This document also provides the more detailed information you requested on the operational plan and field procedures that would be used to give out the incentives. Finally, enclosed is the requested detailed table of the sample loss from the early waves of the Survey of Income and Program Participation (SIPP) through to the 1998 data collection of the SPD.

II. The Revised Proposal

In our continued attempt to slow the rate of attrition for the SPD, we request permission to offer a prepaid incentive of $40 to all Type A noninterview households returning from the 1998 data collection as part of the Type A Follow-up conversion procedure. The Type A noninterviews would receive the incentive in the advance letter, by priority mail, prior to interviewing.1 Type A noninterviews occur when no one is home after repeated attempts, household members are temporarily absent (for example, they are away on vacation for the entire interview period), the household refuses to participate in the survey or other.

We also plan to offer a special $40 incentive to households that could potentially become a Type A noninterview for the first time at the 1999 interview period. These households were not Type As in the 1998 data collection and thus, would not be receiving an incentive in Type A follow-up conversion mentioned above. Households already receiving incentives through the Type A follow-up conversion would not be eligible to receive this special incentive. Details for identifying a household as eligible for this incentive are given in the next section.

It is important to us to prevent any additional noninterview from occurring due to the already sizeable sample loss suffered by this important and unique longitudinal survey.

III. Operations and Field Procedures

A. Incentives for Type A Follow-up Conversion

    Type A noninterviews from the 1998 SPD data collection will be identified for this
    procedure in Headquarters and the Regional Offices (ROs). These cases would receive an
    advance letter containing a $40 incentive, by priority mail, prior to the interviewer's visit.
    The letter will provide information about the incentive and will include the PIN for redeeming
    the amount at an ATM.   The advance letter for these Type As would be sent priority mail to
    increase the likelihood that the letter will be noticed and opened by someone in the sample
    household. Each receiving household would be allowed to cash the incentive regardless of
    the interview outcome (response or nonresponse). Consistent with current practice,
    households who are noninterviews in two consecutive waves will be dropped from the
    sample.

    If the respondent states they did not receive the advanced letter with the prepayment, the Field
    Representative (FR) would be allowed to give an incentive at the door. A system is in place
    to assign and track all debit cards to a case. If a respondent is given a second card, the
    original card is canceled.

B. Incentives for Potential Type As

    Households that were not Type As in the 1998 data collection, but could potentially become a
    Type A would be identified in the field after the 1999 SPD data collection had started.2 FRs,
    as current follow-up procedures require, would notify their ROs of a household's potential
    of becoming a Type A during this data collection period after making several attempts to
    conduct an interview.   The FRs would ask for their RO's assistance by having them send
    either a refusal letter or a no-one-home letter depending on the reason. (These letters ask the
    respondent to reconsider participating in the survey.) The ROs would then send the requested
    letter with a $40 incentive to the household as part of their normal follow-up procedures.
    Again, the requested letter would be sent priority mail to increase the likelihood that the letter
    will be noticed and opened by someone in the sample household and would provide all the
    information about the incentive to redeem the amount at an ATM. The RO would then make
    an attempt to obtain an interview by making a follow-up visit. Each receiving household
    would be allowed to cash the incentive regardless of the interview outcome (response or
    nonresponse).

IV. SPD Sample Loss

There are many reasons why we feel the use of incentives are justified for SPD. These reasons are given in detail in the original proposal document referred to the Executive Summary. One of these is the sizeable household sample loss suffered in SPD. The SPD inherited a 26.6 percent household sample loss rate from the Survey of Income and Program Participation sample. After two waves of SPD, the household sample loss rate is 50 percent. Table 1. offers some of the detailed sample loss information you requested. Similar tables by poverty and nonpoverty will be forthcoming. Previous studies on SIPP sample loss have shown that the sample loss is not uniform. Persons in and near poverty attrit at a higher rate than other persons. Since poverty households are a key target population in the study of welfare reform, there is some concern about nonresponse bias.

Table 1. Overall Household Sample Loss-An Average of the 1992 and 1993 SIPP Panels and SPD

Interview Eligible Initially Assigned Interviewed Refusals (Cumulative) Unlocated Movers (Cumulative Sample Loss

Total Hard Soft

1 43394 43394 39446 3948 3174 774 0 8.8

2 44225 39446 37936 5555 4350 1205 708 14.4

3 45043 40262 37882 5872 4727 1145 1249 16.3

4 45468 38661 37477 6311 5140 1171 1639 18.1

5 45985 38387 36985 6849 5535 1314 2056 20.3

6 46437 37907 36676 7233 5926 1307 2442 21.9

7 46704 37489 36133 7753 6319 1434 2717 23.6

8 47030 36885 35761 8086 6586 1500 3074 25.1

9 47273 36452 35291 8483 6887 1596 3375 26.6

10* 17804 14761 13337 3111 2543 568 1224 26.6

>97 SPD Bridge** 49000 34610 30125 12824 NA NA 5939 41.3

>98 SPD*** 31300 19139 16400 10359 NA NA 4408 50.0

* A 10th interview was only conducted for the 1992 panel for only 3/4 of the sample.

** Only those HHs interviewed in the last wave of the 1992 or 1993 panels were initially assigned to field for the SPD Bridge.

*** Only those HHs interviewed in the Bridge and selected during the subsampling were initially assigned to field for the 1998 SPD.

Timing

We propose giving a $40 incentive to all Type A 1998 SPD cases and all potential Type A 1999 SPD cases for the 1999 SPD data collection which would be in the field at the end of April 1999. We need OMB approval by March 15, 1999 to be able to implement the procedure successfully.

1 Note that all SPD households will be receiving their advance letter by priority mail. There is evidence that using priority mail (in the SIPP) increases cooperation among reluctant respondents.
2 Households eligible for this incentive are those with 1998 SPD outcome of interview, Type B (institutionalized), or Type D (unlocated movers or in US but out of interview range).
4

SPD Working Papers
SPD Publications

Contact: (dsd.survey.program.dyanmics@census.gov)
URL: http://www.spd.census.gov/SPD/