Dynamics of SNAP Participation RFP AG-3198-S-10-0012.pdf

June 2, 2010

Attention: Offerors

The USDA Food and Nutrition Service (FNS), Contract Management Division (CMD) on behalf of the Office of Research and Analysis (ORA), announces this Request for Proposal (RFP) for the attached Performance Work Statement (PWS). Offerors will propose on the entire PWS, all or nothing. It is the Government’s intent to request competitive proposals and to place an order with the Offeror that can provide the required services as outlined in the PWS.

This is a competitive small business set-aside procurement for advisory and assistance services to update and expand our knowledge of the dynamics of participation in the Supplemental Nutrition Assistance Program (SNAP).

Award will be made to the Offeror whose technical proposal and price provide the best value to the Government.

Please provide six (6) hard copies of your technical and four (4) hard copies of your price proposal, including originals (one each), to be received by the office specified below not later than July 13, 2010 by 12:00 noon (EDT). In addition, please provide one CD for the technical factors and one CD for price. Proposals shall be submitted to Belal.Hammad@fns.usda.gov

Hard copies shall be sent to:

USDA Food and Nutrition Service Contract Management Division Attn: Mr. Belal Hammad, Contract Specialist 3101 Park Center Drive, Room 2228 Alexandria, VA 22302

SECTION B – SUPPLIES OR SERVICES AND PRICES/COSTS

B.1. SERVICES NONPERSONAL

The Contractor shall provide all management, supervision, labor and equipment, and shall plan, schedule, coordinate and assure effective performance of all services/tasks as described in the RFP.

The Government anticipates awarding fixed price contract.

PRICING

(a) Insert the labor category, rate, total number of labor hours used to complete CLIN one (1) thru four (4) and the extended total

Overall Percentage discount is: _____ % Base Year Period

Base year period

Labor Category Rate % Discount

Discounted Rate

(b) Insert the total fixed price for each Contract Line Items Numbers (CLIN) and total price.

CLIN # Sub-CLIN

DESCRIPTION

Related Task

Total

FIXED

PRICE

Update SNAP Participation

Dynamics, Describe the Participation Dynamics of Subgroups and Further Explore Circumstances that Influence SNAP Entire and Exits

Insert Price below, in total.

001.01 Expand and Revise the Study Plan

1 $

001.02 Develop and Assess Data Files and

Create Data Base

2 $

001.03 Data Analysis 3 $

001.04 Prepare a Final Report on Dynamics of

SNAP Participation 4 $

CLIN 0001 Total Price $

(c) Travel Costs:

When applicable, travel costs shall be calculated and reimbursed using the per diem rate schedule established by GSA.

SECTION C - Performance Work Statement (PWS) DYNAMICS OF SNAP PARTICIPATION IN THE MID- 2000s

OVERVIEW

The Food and Nutrition Service (FNS) of the U.S. Department of Agriculture (USDA) is sponsoring the latest study in a series that will update and expand our knowledge of the dynamics of participation in the Supplemental Nutrition Assistance Program (SNAP)1.

The study will provide:

• Updated measures of SNAP participation dynamics, including entry rates, spell lengths, exit rates, and reentry rates, which are comparable to previous estimates.

• Expanded examination of individuals’ circumstances that trigger SNAP entries and exits and influence the length of time participants remain on SNAP.

This analysis will be conducted using the 2004 panel of the Survey of Income and Program Participation (SIPP), covering the period between October 2003 and December 2007. This period covers a time when SNAP participation was at high but fairly stable levels, prior to the sharp increase in 2008 and 2009. The study will provide a baseline covering the period before the implementation of the Food, Conservation, and Energy Act of 2008 (2008 Farm Bill) and the American Recovery and Reinvestment Act of 2009 (ARRA), and the onset of the economic recession of 2008-2009. It will update and build upon the analyses performed using the 2001, 1996, 1990-1991, 1987, and 1984 SIPP panels.

BACKGROUND

Supplemental Nutrition Assistance Program (SNAP)

SNAP is the largest of the domestic food and nutrition programs administered by the FNS and its State partners. Assistance is provided in the form of electronic benefits transfer (EBT) cards which can be used to purchase food in authorized food stores.

The program is authorized by Congress, administered nationally by the FNS, and operated by State agencies and their local offices. During fiscal year 2009, the Program served an average of 33.7 million persons in nearly 17 million households each month and delivered a total of $50 billion in benefits. The average monthly SNAP benefit was $276 per household.

SNAP benefits are available to most low-income households with few resources. As a result, the program serves a broad spectrum of needy households, including elderly and disabled individuals, single parents with children, married couples and other adults with

1 In October 1, 2008, the Food Stamp Program changed its name to the Supplemental Nutrition Assistance Program (SNAP). We use SNAP throughout the document, with the exception of when citing past reports that contain the words “Food Stamp” in their titles.

children, and nonelderly/nondisabled individuals. Across the eligible population, there is variation in race, ethnicity, and degree of poverty, frequency and pattern of poverty over time, receipt of benefits from other food and public assistance programs, and employment history. There is also variation in the extent to which eligible individuals participate in SNAP and the length of time they remain in the program. Some of these differences appear to be related to age, working status, and prior SNAP experiences.

SNAP Participation Levels

Participation levels in SNAP are strongly influenced by prevailing economic conditions and program policy. Caseloads, in response to economic and policy changes, have increased and decreased substantially over the last 20 years. After declining steadily during most of the 1980s, the SNAP caseload rose sharply in the early 1990s, rising from 18.4 million in September 1988 to a peak of 28 million in March 1994. This period was followed by gradual declines through 1996 and then more rapid reductions, reaching a low point in July 2000, when fewer than 17 million received benefits. In recent years, participation levels have grown dramatically in response to expansions in eligibility, improved outreach, program simplification in eligibility and certification rules, and a severe economic recession. As of December 2009, 38.98 million persons participated in SNAP.

The 2004 Panel of SIPP covers the period from October 2003 through December 2007.

During this time, average monthly caseloads gradually increased from 23.3 million to

27.6 million2.

Data Source Used to Analyze Participation Spells

Survey of Income and Program Participation (SIPP)

SIPP is a nationally representative, longitudinal survey of the resident non-institutionalized population in the U.S., conducted by the U.S. Census Bureau. Data are collected every four months about each preceding month of the reference period or wave to minimize recall error. The sample in each wave consists of 4 rotation groups, each interviewed in a different month. The core questions, which are repeated at each interview, cover labor force activity, types and amount of income and participation in various benefit programs. Topical Modules are administered every wave covering specific socioeconomic and demographic items. The module that focuses on respondents’ history of employment and program participation is administered in the first wave.

This study will use the 2004 Panel of SIPP. Interviews for Wave 1 were conducted February 2004 through May 2004 for the reference period October 2003 through

2 SNAP participation spiked in the fall of 2005, in the wake of Hurricanes Katrina, Rita, and Wilma. In November, 2005, 29.8 million Americans participated in SNAP, but over 4 million of them were receiving emergency disaster assistance. Most were off the program by January 2006.

January 2004. The number of respondents interviewed was 51,379. The full longitudinal panel is available.

The following describes the SIPP data and briefly discusses studies of particular relevance for the proposed new SNAP research.

Previous Research on SNAP Participation Dynamics

Generally speaking, studies of SNAP participation dynamics analyze data that follow individuals over time enabling researchers to measure how frequently people with different characteristics participate in SNAP and for how long.

Since 1990, a series of studies using SIPP panels have tracked individuals’ SNAP participation over 30 to 48 months, depending on the panel. The purpose of these studies is to answer questions about how many individuals enter SNAP over the panel period, how long they stay on the program after entering, whether they reenter the program after exiting, and their reasons for these decisions. These studies examined whether caseload increases or declines are fueled by changes in entry rates among eligible persons and/or longer or shorter durations on the program. They also provide descriptive information about the composition of SNAP caseload at a point of time, answering questions about who receives SNAPs for short and long periods of time, and whether their participation is continuous or periodic. Multivariate analyses answered questions about which factors seemed to have the most influence on participants’ movement on and off SNAP.

Dynamics of the Food Stamp Program as Reported in the Survey of Income and Program Participation

This study was the first to use SIPP data to provide a comprehensive picture of participation dynamics among persons who entered SNAP in the early and middle 1980s. Burstein (1993) used the core information from the 1984 panel and found that most persons entering the program rely on SNAP for short periods of time. Burstein found that the median spell length for SNAP entrants was only 6 months, that two-thirds of spells ended within a year and that almost 40 percent of those who exited returned to SNAP within a year. Burstein also documented that income changes, particularly earnings, preceded most entries into and exits from SNAP.

The methodology employed in the descriptive analysis entailed numerous decisions.

They included subgroup classifications, treatment of missing months, definitions of relevant triggers, selection of appropriate thresholds for income change and the length of the window for evaluating trigger events.

Characteristics of Long Term Participants in the Food Stamp Program

Murphy and Harrell (1992) used data from the 1987 panel of SIPP to categorize SNAP participants into four participant groups: participants with a single spell that was short-term, medium-term, or long-term, and participants with multiple spells (i.e., cyclers).

The study found that by the end of the 28th month of the panel, a cross-section of participants was dominated by long-term participants (59 percent) and cyclers (27 percent).

The Dynamics of Food Stamp Program Participation in the Early 1990s

Gleason et al. (1998) used data from the 1990 and 1991 panels of SIPP to describe the experiences of SNAP participants at a time when caseloads were increasing. This work updated and augmented the research presented in Burstein. The study produced descriptive measures of spells for both an entry cohort sample and a cross-section sample. The latter was developed with information collected in the Recipiency History and Employment History topical module. The incidence of multiple spells during the SIPP period was captured by measuring the total amount of time individuals participated in SNAP either continuously or intermittently. The study also included a multivariate analysis of spell duration and an exploratory analysis of the events that trigger SNAP entry, controlling for long-term, more permanent participant characteristics

Gleason et al. found that the median spell length among SNAP entrants had lengthened to 9 months in the early 1990s and that less than one-third of entrants remained on the program for 2 or more years. Fewer than one in five remained on SNAP for five or more years. The cross-sectional analysis confirmed the Murphy and Harrell findings that there is greater dependence on SNAP among participants at a given point in time. In a given month, about 10 percent of participants were in the middle of a spell of a year or less, over three-fourths were in the middle of a spell for two or more years, and over 60 percent were in the middle of a spell of five or more years. There was also evidence of higher reentry rates within the first two years after exiting. The median total time on SNAP during the 32 month period of the 1991 panel was 20 months.

The study also confirmed Burstein findings that income changes trigger most movement into and out of SNAP although there was also some evidence that a combination of falling income and a change in household composition also commonly led to a SNAP entry.

The study included a multivariate analysis on the determinants of SNAP participation spell length. Those with incomes below poverty and those with longer periods of not working have the longest SNAP participation spells. Persons receiving cash welfare, all else equal, when they enter SNAP tend to stay on longer than those not receiving cash welfare. Similar factors influence reentry rates. Among household factors, households with fewer adults and more children have longer spells.

Gleason et al. also examined questions about the relationship between the long-term and prior month circumstances of participants in explaining SNAP entries. They conducted multivariate analyses on a sample of adults age 18 to 49 from the 1991 SIPP panel to test various combinations of current and past circumstances. The preliminary results suggest that for the most part, average variables (defined as covering the past two years, such as average two-year earnings) are not significant predictors of SNAP entry but that prior month trigger variables were powerful predictors when they represented a deviation from the past. For example, the effect of being unemployed last month has a greater effect on entry for those who had high average employment levels than low-average employment levels. This analysis did not, however, look at the full range of past experiences and did not control for past SNAP receipt. It also did not look at the effect of these factors on duration of SNAP spells.

In terms of overall experiences, Gleason et al. concluded that rising caseloads in the mid-1980s and early 1990s were driven primarily by an increase in the length of participation spells.

Food Stamp Program Entry and Exit: An Analysis of Participation Trends in the 1990s

Cody et al. (2005) primarily examined patterns in SNAP entry and exit rates in order to determine how these patterns contributed to caseload trends in the 1990s. The study also updated earlier estimates of participation spells for new entrants and all participants in any given month. These analyses were conducted for subgroups and three periods representing different stages of SNAP caseload growth cycles. These were the period of growth from 1990 to 1993, the period of some caseload decline between 1993 and 1996 and the period of sharp caseload decline from 1996 to 1999.

All measures were estimated using SIPP panels for 1990 to 1996, and SNAP Quality Control data from 1990 to 2002. Although QC data are not longitudinal, information on SNAP entry dates were used to estimate time on SNAP and link active cases across months. Some of the key findings from the study were:

• More than half of new entrants exited the program within 8 months, and about two-thirds of new entrants exited within one year in the 1990s.

• In March 1996, between one-third and one-half of participants were in the middle of spells longer than four and a half years. (This represents a substantial decline from earlier estimates.) Long-term spells, however, became shorter over the 1990s. According to SIPP data, the top quartile of cases had spell lengths of 26 months in the early 1990s and 16 months in the late 1990s.

• Caseload changes during the 1990s were driven by both changes in the rate that people entered the program and the length of time that people participated.

Much of the caseload growth of the early 1990s was caused by increasing replacement rates, although lengthening spells also contributed to the growth.

Much of the caseload decline of the late 1990s was caused by substantially shorter durations in long-term participation spells.

Because Cody et al. based much of his analysis on SNAP QC data, which does not contain any data on non-participants; the analysis did not examine SNAP entry as a proportion of the population not on SNAP. Rather, this analysis looked at the number of new entries as a proportion of the caseload from the prior month, terming this measure the “replacement rate”. This measure is not comparable with the entry rate as a proportion of the population not on SNAP used in prior studies.

Cody et al. also identified some concerns about the 1996 SIPP data that may affect comparisons with estimates of participation measures based on earlier SIPP panels.

The study found three key inconsistencies: a change in the proportion of adults among SNAP participants, changes in SNAP participation volatility, and changes in the number of new SNAP households with earnings. The report suggests caution using the 1996 panel without some adjustments. These problems were addressed in the next analysis, using the 2001 SIPP panel.

Dynamics of Food Stamp Program Participation, 2001-2003

Cody et al. used data from the 2001 SIPP panel to examine individuals’ patterns of participation in SNAP. The seven research questions that were addressed included:

1. What factors led individuals to enter the program?

2. How long did individuals tend to participate?

3. What factors led individuals to leave the program?

4. How frequent was program re-entry?

5. How much did individuals rely on the program over time?

6. How did participation patterns vary by subgroup?

7. How did participation patterns change since the early, middle, and late 1990s?

The study included a descriptive analysis of the characteristics of participation spells observed from 2001 through 2003, estimated rates of entry onto the program, the duration of participation spells, the likelihood of re-entering the program within the panel period, and events associated with entry onto and exit from the program.

The second stage of the study included a multivariate analysis of the factors associated with entry, spell duration, and re-entry.

Some of the key findings included the following:

• Participants tended to enter and leave SNAP rather frequently; half of all those who entered between 2001 and 2003 left within 8 months and 61 percent left within a year.

• In any specific month, most participants had been on the program for longer periods; half were on for four years or more and many for more than 8 years.

• SNAP served substantially more people over the course of a year than it did in an average month; the turnover ratio during this period was 1.5.

• Most individuals came onto the program because of a substantial change in household income, usually earnings. However, most entrants had multiple changes in circumstances including changes in household composition.

• Households with more volatile employment histories were less likely to enter SNAP than those with more stable histories and a recent substantial change in earnings.

Understanding the Determinants of Supplemental Nutrition Assistance Program Participation

Burstein, et al. used data from the 2001 and 2004 SIPP panels to develop two models that would identify factors that contribute to households’ decisions to participate in SNAP. One of the models was a standard model, which included measures of what are generally recognized to be key factors in SNAP participation. These variables included household income, presence of children, presence of earnings, education, age, race, ethnicity, marital status of the reference person, and the local unemployment rate. The second model used an expanded set of factors that were suggested in previous studies as being participation predictors. It included more detailed measures of household income and resources (such as whether income is below 50 percent of poverty as opposed to simply below poverty); more detailed data on dependents (such as finer breakdowns of children by age and the number of elderly adults); household members in poor health; health care expenses; child support payments; recent moves; labor market participation; current or past receipt of Temporary Assistance For Needy Families (TANF) or other variables; receipt of other income; and certain State policies.

The study found that the expanded model predicted participation better than the standard model. Variables that were strongly related to SNAP participation included the use of biometric technology, such as fingerprinting (negative correlation to participation);

short certification periods (negative); the number of children age 12 and under (positive); the State unemployment rate (positive); income relative to poverty (as income rose, participation fell); bad health (positive); education level of the adults (negative);

whether the household head was married (negative); receipt of TANF in the past or present (positive); the presence of earnings (negative); and the presence of means-tested cash and non-cash benefits (positive).

STUDY OBJECTIVES AND RESEARCH QUESTIONS

The purpose of this study is to advance our knowledge of SNAP participation dynamics by:

(1) Developing measures of recent SNAP participation dynamics, including entry rates, spell lengths, exit rates, and reentry rates, which are comparable to previous estimates.

(2) Describe the participation dynamics of subgroups.

(3) Further analyzing the influence of a broader range of factors, recent and past, on SNAP participation decisions.

Objective 1: Update SNAP Participation Dynamics

Earlier work by Burstein, Gleason et al. and Cody et al. developed estimates of SNAP participation spell lengths by subgroups of individuals; estimates of rates of SNAP entry, exit and reentry; estimates of the importance of individuals’ circumstances on SNAP entry and exit; and estimates of measures of participation over time. Replication of this research using the 2004 SIPP panel set will be guided by the following research questions:

• How long are spell lengths for entrants? How do these vary among different SNAP subgroups?

• What proportion of participants who exit SNAP return to the program within 6 months, within a year, within 2 years? What is the median time between SNAP participation spells? How do re-entry patterns vary among different subgroups?

• How long are spell lengths for participants when viewed at a specific point in time? What proportion of the caseload has single short spells (1-8 months), single medium-term spells (9-23 months), single long spells (24+ months), or more than one spell during the SIPP panel period?

• What are entry and exit rates in the period covered by the 2004 SIPP panel for all individuals “at risk” (below 200% of poverty and not on SNAP) of entries and exits and for SNAP subgroups? Have these changed since the early 2000s?

• How many individuals leave SNAP in the period covered by the 2004 SIPP panel for reasons that are not related to improved financial circumstances or reduced need as measured in SIPP? What are the spell lengths of nonparticipation among these leavers, compared with others leaving the program?

• How sensitive are the findings on program entry to the definition of the population at risk (for example, below 100% of poverty or below 300% of poverty)?

• What participant characteristics distinguish those who have longer spells, frequent spells, or spend a significant proportion of the panel time on SNAP from those who only use the program for a short single spell, or a small proportion of the panel time?

• What dynamics explain the participation growth that occurred in the mid- 2000s?

What are the replacement and exit rates in this period?

• What is the turnover rate (the ratio of all participants ever on SNAP during the year over the average monthly number of participants) for SNAP participants in each year covered by the 2004 SIPP panel? Did the turnover rate change between the first few waves of the panel, which was during a period of economic growth, and the later waves of the panel, when the economy was starting to weaken?

• How do the answers to the above research questions compare to the findings in the studies for the earlier periods?

• Based on the assessment of the 2004 SIPP panel data, what are the patterns of one- and two- month gaps in benefits that are attributed to churning? These gaps are usually disregarded in the preparation of analysis files. Does the incidence of churning correlate to any participant characteristics (e.g., new entrant versus reentrant), spell length, benefit size, administrative milestones (e.g. recertification), or State practices (e.g. caseworker assists in filling in application and recording changes versus online application and call centers)?

Objective 2: Describe the Participation Dynamics of Subgroups

Past studies on SNAP participation dynamics have examined entry, exit, recidivism, and spell length for all SNAP participants, with analyses showing variation in each of these measures among different subgroups. Another approach would be to examine the dynamics of each subgroup, and explain how their patterns relate to the characteristics of the overall SNAP caseload from different perspectives. For example, prior to welfare reform, single parents comprised a relatively small share of the population of persons with income below 300 percent of poverty, but because they were much more likely to come on the program and because they had longer participation stays, they represented a much larger share of the SNAP caseload. Is that still true today? How do the dynamics of adults and children in single parent households compare to those in households with married parents or unrelated adults? How do the differences between these dynamics explain the composition of the SNAP caseload? What insights about the characteristics of the SNAP caseload a can be told by examining the dynamics of other groups and comparing them to one another?

Important subgroups to examine this way could include, if sample size permits:

• Single parents;

• Children of single parents;

• Married adults with children;

• Children of married adults;

• Elderly adults;

• Non-elderly disabled adults;

• Non-elderly nondisabled childless adults;

• Noncitizens;

• People living in households with earnings;

• People living in households with Social Security income;

• People living in households with Supplemental Security income;

• People living in households with TANF income; and

• People living in households with zero income.

This objective requires taking what was learned from Objective One and reframing it in terms of the size and dynamics of the aforementioned subgroups.

Objective 3: Further Explore Circumstances that Influence SNAP Entries and Exits.

Cody et al. presented findings from a multivariate analysis of possible determinants of SNAP entry. Burstein et al. expanded on these findings and added additional possible determinants, but only examined these variables in terms of predicting participation.

This objective shall explore these additional variables further, looking at any relationships between them and spell length, exit rates, and recidivism rates. This objective shall also explore potential determinants, such as citizenship status and country of birth for the foreign-born or the presence of a broad-based categorical eligibility policy in a State. Specifically, Objective Three will address the following research questions:

• Controlling for the characteristics of individuals, what circumstances are most prevalent just prior to SNAP entries and exits? How much impact do factors such as employment, marital status, annual income, prior participation in SNAP, and similar circumstances affect SNAP entries and exits?

• What trigger events precede SNAP entry in the study period? What trigger events precede reentry or what are the characteristics of participants that lead to reentry? Which trigger events are the most powerful in predicting SNAP entries and reentries?

• How does stability or instability of individuals’ circumstances affect SNAP entries and exits? Do these findings vary by subgroup?

• Do the additional variables identified by Burstein, et al. add to our knowledge on what contributes to entries, exits and spell length?

• What is more powerful in explaining the dynamics of key SNAP subgroups:

specific circumstances such as loss of income or the onset of poor health or the number of such difficulties? If multiple problems strongly influence participation behavior, does it matter what they are?

• Can these models explain spell durations or total time on SNAP?

SCOPE OF WORK

Task 1 – Expand and Revise the Study Plan

The Contractor shall make minor revisions and further elaborate on the study plan included in the Technical Proposal. The revisions will incorporate any new information obtained on issues such as, but not limited to:

• Content of relevant data sources.

• Shifting relative priorities among study objectives.

• More specific plans for data analysis.

The Contractor shall prepare a draft of the revised study plan for FNS review. The revised study plan shall incorporate and expand on the study plan included in the Technical Proposal. The contractor shall prepare the revised study plan as a complete stand-alone document. After FNS review, the contractor shall make any necessary changes and submit a final revised study plan to FNS.

Task 2 - Develop and Assess Data Files and Create Data Base

The Contractor shall obtain the 2004 SIPP longitudinal file, the Recipiency History and Employment History Module and any additional SIPP modules or data sets discussed in the Study Plan. The contractor shall then link the SIPP module(s) to the SIPP longitudinal file to create a data base. This file is expected to contain data on SNAP participation spells, participation in other Federal programs (e.g., National School Lunch Program, Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), TANF, and Supplemental Security Income), characteristics of SNAP participants, and other information needed to address the study objectives.

The Contractor shall perform a preliminary analysis of the 2004 SIPP data set. The analysis will determine the extent of any inconsistencies between the longitudinal data and the information in earlier SIPP panels and administrative data for SNAP (for example, the extent of underreporting of SNAP participation in SIPP). This preliminary analysis will address, at a minimum, the concerns about the 1996 and 2001 SIPP panels that are discussed in Cody et.al. The Contractor shall provide information on the incidence of 1-month and 2-month gaps in benefits that are attributed to churning and are disregarded for the analysis of SNAP spells. The Contractor will provide information on the extent of one- and two-month gaps in participation overall and for different subgroups; whether the incidence of these gaps varies by spell lengths, time in the spell, benefit size, demographic characteristics of participants, administrative action, or State practices. For example does churning occur more or less frequently in States with “modernized” certification systems versus States using the traditional caseworker in-person visit model? Is there any way to quantify the SNAP benefit amounts not received by those who experience one or two-month gaps in participation?

The Contractor shall prepare and submit to FNS a memorandum describing the approach and results of the assessment of the 2004 SIPP longitudinal file including the information on one and two-month benefit gaps If the assessment reveals data problems, the contractor shall describe the consequences of the problems for meeting study objectives and present options for addressing the problems.

The Contractor shall also assess the reliability of the retrospective SNAP recipiency history data contained in the 2004 SIPP Recipiency History Module and data in other SIPP modules to be used in this study. This preliminary analysis will address, at a minimum, the concerns discussed in Gleason et al. and Cody et al. If problems are detected, the contractor shall detail an approach for addressing these, including the creation of algorithms or other data adjustments.

The Contractor shall prepare and submit to FNS as part of the memorandum described above the approach and results of the assessment of SNAP recipiency history data and any other proposed SIPP module. If the assessment reveals problems, the contractor shall describe the consequences of the problems on the study objectives and present options for addressing the problems. As part of this memorandum, the contractor shall identify any other data sources proposed to address study objectives.

The Contractor shall submit a memorandum to FNS as specified in the Schedule of Deliverables. Following submission of the memorandum and consultation with FNS the contractor shall complete construction of the data base for this study. Final content and structure shall be consistent with the Study Plan approved under Task 2 and the subsequent Task 3 discussions with FNS.

Task 3 - Perform Data Analysis

The Contractor shall analyze the data collected in accordance with the overall Study Plan and the specific data analysis approach contained in the Study Plan.

The Contractor shall consult with FNS regarding any challenges that impede the goal of successfully addressing each objective and research question. These include any issues that affect the comparability of updated measures of SNAP participation dynamics with the measures developed in earlier studies.

The Contractor shall submit a set of tables to FNS as specified in the Schedule of Deliverables. Upon review by FNS, the contractor shall incorporate any changes into the draft Final report, as specified in the Statement of Work, Task 5.

Task 4 - Prepare a Final Report on Dynamics of SNAP Participation

The Contractor shall prepare a report that addresses all the study objectives and associated research questions. The main report will present all findings but will limit the discussion of detailed analyses to those pertaining to Objectives One and Two. A full discussion of the multivariate analyses that address Objective Three will be presented in a technical appendix or a separate companion volume. At a minimum, the following sections shall be included in the final report:

• An executive summary highlighting the report's major findings.

• An introduction and background section describing relevant issues and research objectives.

• A description of the study data, analytic methods and limitations

• A detailed discussion of any significant results.

• A discussion linking, if possible, the results from analyses across time periods.

• The contractor's conclusions.

• A technical appendix or separate volume to present the detailed results of the multivariate analyses.

• Any additional technical appendixes necessary to fully document the research method and results.

The Contractor shall prepare two drafts and a final report as required in the deliverable schedule.

The Contractor shall provide an electronic copy of the draft presentation to FNS for review and comment. After making any revisions required by FNS, the contractor shall present the results of the project to FNS staff. This presentation shall be held at FNS Headquarters in Alexandria, Virginia. The contractor shall provide an electronic copy of the final presentation to FNS.

The Contractor shall supply FNS with a set of CD-ROMs containing the final SIPP data set in SAS format and the SAS analysis files used to produce the findings.

Concurrently, the contractor shall submit a codebook that describes the data on the CD- ROM, and an updated Technical Description. Documentation shall include: file structure (data set name, record format, record length, block size, and number of records); code book (record layout including variable names, variable format, variable labels and missing values); and formulation of any calculated values.

SECTION D-PACKAGING AND MARKING

D.1 PACKING FOR DOMESTIC SHIPMENT (AGAR 452.247-72) (FEB 1988)

Material shall be packed for shipment in such a manner that will insure acceptance by common carriers and safe delivery at destination. Containers and closures shall comply with the Department of Transportation regulations, Uniform Freight Classification Rules, or regulations of other carriers as applicable to the mode of transportation.

SECTION E-INSPECTION AND ACCEPTANCE

E.1 INSPECTION AND ACCEPTANCE (AGAR 452.246-70)(FEB 1988)

(a) The Contracting Officer or the Contracting Officer's duly authorized representative will inspect and accept the supplies and/or services to be provided under this contract.

(b) Inspection and acceptance will be performed at:

Project Office:

USDA Food and Nutrition Service Family Programs Analysis Branch Office of Research and Analysis Attn: Jenny Laster Genser 3101 Park Center Drive, Room 1014

Tel: 703-305-2559

E.2 SCOPE OF INSPECTION

All deliverables will be inspected for content, completeness, accuracy, and conformance to task order requirements by the COR. Inspection may include validation of information or software through the use of automated tools and/or testing of the deliverables, as specified in the task order.

E.3 BASIS OF ACCEPTANCE

The basis for acceptance shall be compliance with the requirements set forth in the performance work statement, the acceptance measures in the schedule of deliverables, the contractor's proposal and other terms and conditions of the contract. Deliverable items rejected shall be corrected in accordance with the applicable clauses.

The Government shall have up to fifteen (15) business days upon receipt to review each deliverable and provide oral and written comments. The Contractor shall review and incorporate comments or implement directed changes, after discussion or clarification from the Project Manager and submit a final version of the deliverable no later than ten

(10) business days thereafter.

NOTE All stated days for delivery and government review, unless denoted otherwise, are stated in business days.

SECTION F-DELIVERIES OR PERFORMANCE

F.1 IDENTIFICATION OF CONTRACT DELIVERABLES

Unless otherwise specified by the Contracting Officer in writing, all documents prepared and submitted by the contractor to the Government under this contract shall include the following information on the cover page of each document:

(a) Name and business address of the contractor;

(b) Contract number

(c) Contracting Officer’s Representative’s Name; and

(d) Date of delivery order

F.2 PLACE OF DELIVERY-FOB DESTINATION

All reports and/or deliverables identified under the contract shall be delivered, and all transportation charges paid by the Contractor, in accordance with the FOB Destination clause (FAR 52-247-35) as follow:

Project Office:

USDA/Food and Nutrition Service Office of Research and Analysis Attn: Jenny Genser 3101 Park Center Drive, Room 1014 Alexandria, VA 22302

Progress Reports and a copy of all invoices shall be delivered electronically to the Contracting Office and the Project Office at the addresses shown below:

jenny.genser@fns.usda.gov Belal.hammad@fns.usda.gov

F.3 TIME OF DELIVERY

The Government requires delivery to be made in accordance with Schedule of Deliverable (TBD).

F.4 PERIOD OF PERFORMANCE

The period of performance will be 12 months from the date of award.

F.5 Monthly Reports

The contractor shall provide FNS with monthly reports on the progress of the study.

Each progress report shall include the following:

• Cover page with contract number, title, period of performance covered and number in the sequence of reports.

• Brief description of the purpose and scope of the study.

• Description of activities, by task, that were carried out during the reporting period.

• Description of the activities planned for the next reporting period;

• Discussion of problems and delays encountered during the reporting period and steps taken or proposed to resolve them.

• Discussion of anticipated problems and recommendation for avoiding or resolving them.

• Schedule identifying for each task and deliverable the dates on which work started, was expected to end, and actually ended.

One copy of the monthly report shall be submitted directly to the FNS Contracting Officer, and one copy to the Contracting Officer’s Technical Representative. These reports shall be submitted no later than the 20th working day of the following month.

F.6 All Other Deliverables

All required deliverables (with the exception of the monthly progress report) shall be delivered to the Contracting Officer’s Representative as follow: See Schedule of Deliverable (Attachment II).

SECTION G-CONTRACT ADMINISTRATION

G.1 POST AWARD CONFERENCE (AGAR 452.215-73)(NOV 1996)

A post award conference with the successful offeror is required. It will be scheduled within 10 days after the date of contract award.

The contractor shall meet with FNS staff in Alexandria, Virginia during the second week after the contract award to review, discuss and clarify the study objectives delineated above, program management plans, and schedule. At a minimum, the following items shall be addressed:

• A protocol for all communications between FNS and contractor staff, including financial reporting, periodic progress reports, and procedures for product approval;

• Clarification of project schedule, contents of deliverables, and staffing assignments as needed; and

• Strategies for handling potential problems, such as changes in variable coding or other data irregularities.

The contractor shall submit an agenda for the orientation meeting one week after the contract award for review by FNS.

The contractor shall submit a memorandum to FNS within 2 weeks after the orientation meeting documenting the issues discussed, the decisions made, and any outstanding matters.

G.2 CONTRACT/TASK ORDER ADMINISTRATION

(a) The Contracting Officer is the only person authorized to approve changes in any of the requirements of the task order. This authority remains solely with the Contracting Officer. In the event the contractor effects any change at the direction of any person other than the Contracting Officer, the changes will be considered to have been made without authority and no adjustment will be made in the contract price to cover any increase in costs incurred as a result thereof.

(b) Only the Contracting Officer, Contract Specialist and Contracting Officer’s Representative (COR) have the authority to contact the contractor to discuss costs and/or technical requirements of the contract.

(c) Contact from other government employees could result in unauthorized commitments. Government representatives who are working on, or asked to work on, a particular contractual issue must not contact the contractor directly, but should coordinate the activity with the Contracting Officer or COR. The COR must be present at all procurement related contacts and meetings with the contractor. Depending on the nature of the meeting, the COR must also give advance notification to the Contracting Officer or Contract Specialist. All contacts with the contractor must be documented and retained with the COR’s contract correspondence file. As deemed appropriate, one copy must be forwarded to the Contracting Officer. Examples of documented contacts are providing technical direction, work plan instructions, not accepting deliverable, potential changes to the Performance Based Work Statement or delivery schedule, costs and subcontractor or consultant recommendations.

(d) The contractor shall designate one individual to be contacted during the period of the contract for prompt contract administration:

Name:* Title:* Phone:* Email:* *to be completed by offeror

(e) Contracting Officer (CO):

To be provided under separate cover.

(f) Contracting Officer’s Representative (COR):

Jenny Laster Genser

Family Programs Analysis Branch Office of Research and Analysis 3101 Park Center Drive, Room 1014

Tel: 703-305-2559

G.3 CONTRACTING OFFICER’S REPRESENTATIVE (COR):

Definitions:

“Contracting Officer’s Representative” – The Contracting Officer’s Representative (COR) is designated by the Contracting Officer, in writing, and is contract order specific.

The COR is the technical representative for all technical requirements of the specific contract.

“Contracting Officer” – The Contracting Officer (CO) is the exclusive agent of the Government with the authority to enter into, and administer contracts. Thus, the CO has the responsibility to see that all requirements of the law and regulation are followed.

However, as the CO’s representative the COR is delegated the authority to monitor the technical effort being performed under the contract. The CO is the only individual with authority to enter into, administer, or terminate contracts. The CO ensures performance of all necessary actions for effecting contracting, ensuring compliance with the terms of the contract and safeguarding the interests of the United States in its contractual relationships.

Limits on the Authority of the COR:

The authority of the COR is limited. The CO may not delegate the following authority to:

1. Award, agree to, or execute any contract, task order, or contract modification;

2. Obligate, in any way, the payment of money by the Government;

3. Take any action which may have an impact on the contract or task order schedules, funds, or the scope of work/task description/specifications;

4. Make the final decision on any matter that would be subject to appeal under the Disputes clause; and

5. Terminate, for any case, the contractor’s right to proceed.

Responsibilities:

The COR is authorized to take any or all actions as specified below which could lawfully be taken by the CO, except any action specifically prohibited by the terms of the contract.

1. Maintain an arms-length relationship with the contractor in the interest of procurement integrity as well as sound contract management.

2. Keep the CO fully informed of any technical or contractual difficulties encountered during performance. The COR should also advise the CO of any potential problem areas under the contract.

3. Assure that the contractor performs the contract technical requirements in accordance with its terms, conditions, and specifications.

4. Inform the contractor of failures to comply with the technical requirements of thiscontract, and inform the CO of any failures to do so, particularly if the contractor does not make corrections.

5. Coordinate site entry for contractor personnel, if applicable.

6. Ensure that Government furnished property, if any, is available when required, and report any accountable property to the appropriate property personnel.

7. Ensure that all required items, documentation, data, and/or reports are submitted to the COR as required by the contract. If additional time is required by the contractor, the contractor shall submit a formal request for a time extension to the CO through the COR. The COR shall indicate concurrence or state the reasons for non-concurrence, and forward the request to the CO for finalization.

8. Evaluate proposals for and participate in negotiation of changes, modifications and claims at the request of the CO.

9. Review vouchers for cost-reimbursement type work and recommend approval by the CO if the contractor’s costs are consistent with the negotiated amounts and progress is satisfactory and commensurate with the rate of expenditure (if applicable).

10. Review and approve invoices for fixed-price deliverables to ensure receipt of the goods and/or services.

11. Process all invoices and vouchers in a timely manner in accordance with the Prompt Payment Act.

12. Document actions taken and decisions made as the COR, and maintain adequate records to sufficiently describe the performance of the duties as COR during the life of the contract.

13. Provide the CO with a copy of any correspondence between the COR and the contractor.

14. If the contract is for construction or services and you visit the site where work is being performed, check to see that the Department of Labor and Equal Employment Opportunity posters and applicable wage determination rates are posted in full view of employees.

15. Perform all inspections necessary in with the contract and require the contractor to correct any deficiencies that are found; recommend acceptance for all deliverables on behalf of the Government; record the inspections and acceptance and insure that distribution of approvals/disapprovals are made to the contractor and the CO. Review all contract deliverables and provide inspection and acceptance documentation to the CO. The COR has the responsibility to track and retain all contract deliverables. Assist the CO with contract closeout activities as requested.

16. Submit reports on contractor performance as required.

17. Evaluate contractor performance in accordance with Federal Acquisition Regulations (FAR) Part 42.15 and acquisition office procedures.

18. During the pre-award phase of this contract, the COR is responsible for performing the following duties as it applies to the contract/task order:

(a) Provide the Contracting Officer with the appropriate contractor personnel security screening requirements and background investigation (BI) requirements for obtaining services of non-federal employees under the resultant contract or task order.

(b) Determine the risk level for each contractor position. The risk level must be based on an overall assessment of the damage that an untrustworthy contractor could cause to the efficiency…