Return to data section
To print: Select File --> Print from your browser's menu.
............................................................................
Note: any redistribution of this report must provide credit to the authors
and sponsoring organizations.
THE EFFECTS OF WELFARE REFORM
ON PRIVATE EMERGENCY ASSISTANCE
Interim Report
For The Local Investment Commission
3100 Broadway, Suite 226
Kansas City, Missouri 64111
MRI Project No. 1445-02
March 12, 1998
Preface
----------------------------------------------------------------------------
Midwest Research Institute (MRI) is pleased to present to the
Local Investment Commission (LINC) an interim report on the potential
effects of welfare reform on privately provided emergency services
in Jackson County, Missouri. This project is being jointly funded by LINC,
through a grant from the Missouri State Department of Social Services,
and a Kimball Research Award from MRI.
The purpose of this study is to investigate the potential impact of
welfare reform on the demand at food pantries, community emergency shelters,
privately provided utility assistance, and other emergency social services.
Information for this study is being obtained from the Missouri Department
of Social Services and from the Mid America Assistance Coalition (MAAC).
The authors of this report are Nancy Dunton and Richard Speidel of
MRI’s Center for Regional Development. Wanda Brandenburg and
Brian Rosson assisted in the analysis.
This interim report provides LINC with some of the first findings
on the joint utilization of both public assistance and private emergency
assistance in Jackson, Clay, and Platte counties. The final report will
present additional information on the joint usage of food stamps and
private assistance, as well as a longitudinal analysis of the relationship
between these services. The longitudinal analysis will provide information
on the degree to which families rely on private assistance as they leave
public assistance behind. This information will help Kansas City, through
LINC and MAAC, plan to meet the changing emergency service needs of the
low-income population.
Sincerely,
MIDWEST RESEARCH INSTITUTE
Nancy Dunton, Ph.D.
Principal Social Scientist
Approved:
James L. Spigarelli, Ph.D.
Executive Vice President, MRI
Director, Kansas City Operations
March 12, 1998
Contents
----------------------------------------------------------------------------
Preface
Introduction
Interim Report Contents
Methodology
Preliminary Findings
Next Steps
Appendix -- Procedures for Protecting Confidentiality
Introduction
----------------------------------------------------------------------------
Public concern about the welfare system nationwide has been evident for
more than a decade. This concern resulted in the implementation in many
states of waivers to the Aid to Families with Dependent Children (AFDC)
program and the passage of federal and state welfare reform legislation.
Specific welfare reform activities in Missouri are described below.
The intent of these reforms is to improve the economic self-sufficiency
of persons and help them make the transition from welfare to work. Another
intended benefit of reform is to reduce the cost of the welfare system by
reducing the welfare rolls and enabling former recipients to become taxpayers.
As with any major change in a social institution, there are legitimate
concerns about whether the effects of the change will match the intention of
the reform.
The focus of this study is to identify the amount and pattern of
private assistance used by public assistance recipients, as they leave
the public caseload. This analysis is made feasible by the unique data set,
maintained by the Mid America Assistance Coalition, on private emergency
assistance services in the Kansas City metropolitan area. Before examining
the transition question, however, we must look first at the number of
households using both public and private assistance and at whether this
pattern of joint recipiency has remained stable in the last three years.
We must understand patterns of concurrent recipiency, as defined by
households that use both public and private assistance within the same
calendar quarter, if we are to interpret patterns of transition from
welfare to work.
Before the question of the effect of welfare reform on private
assistance can be assessed, we must first establish baseline information
on the level and pattern of joint recipiency. Therefore, the first
objectives of this study are to:
1. Assess the level of joint usage of public and private
assistance
2. Estimate the number and value of private assistance used by
joint users
3. Describe the income package of joint users
4. Identify neighborhoods with high rates of joint usage
Interim Report Contents
This interim report contains:
Review of the methodological approach and data sources used in
the data analysis
Description of preliminary findings
Discussion of next steps
Methodology
----------------------------------------------------------------------------
The study used data furnished by the Missouri Department of Social
Services (DSS) and by the Mid America Assistance Coalition (MAAC). The
data furnished by DSS included Income Maintenance (IM) monthly extracts
and Food Stamp monthly extracts for the thirty-six month period,
November 1994 through October 1997. Each monthly extract included
information concerning all persons in the counties of Jackson, Clay,
and Platte who were on active cases and receiving some form of
government assistance as of the end of each month.(1) The data
furnished by MAAC included a listing of all services delivered
by member agencies through approximately the same thirty-six month
period, November 1994 through October 1997. The MAAC data included
information detailing the dollar value of the delivered service, the
delivering agency, the recipient's name and social security number
(SSN), and the date the assistance was provided.
Due to the personal and confidential nature of the information
being examined in the study, assuring confidentiality of the data
itself was the first project activity before any data files were
made accessible. MRI's Institutional Review Board (IRB) reviewed and
approved a series of procedures to ensure the confidentiality of
the data. A draft of these procedures is attached as an appendix.
Additional security steps were implemented since the initial portion
of the analysis had to be performed on the mainframe computer at the
University of Missouri-Columbia (UMC).
The analysis of joint receipt of both public and private assistance
services entailed extracting from the IM monthly files, all SSNs of
individuals who received cash assistance at any point within the
selected period. In cases where a portion of the analysis concerned
only individuals or households which received Temporary Assistance to
Needy Families (TANF) cash grants, such recipients were identified by
additional descriptive data on the type of assistance received, also
contained within the IM monthly file extracts. Portions of the analysis
also incorporated individuals who received Food Stamps at any point
within the selected time period by appending the SSNs of such individuals
to the group of public cash grant assistance recipients extracted from
the IM monthly files. If an individual in the Food Stamp extracts had
already been selected by appearing in the group of IM assistance
recipients, the duplication was deleted.
Once a set of unique SSNs was compiled for the selected period
and chosen public assistance grant type recipients, the set was matched
by SSN to records in the MAAC data set for persons with service dates
falling within the same time period. The set of matched records then
had certain details concerning the provision of private assistance
compiled into summary statistics. For example, the number of MAAC
services provided to individuals who also received TANF cash grants
was totaled for each providing agency. Similarly, the total dollar value
of assistance distributed to TANF cash recipients by MAAC agency was
calculated. In most instances, the dollar value of MAAC assistance
was totaled over the selected time period for each recipient.
Once a matched set of SSNs was identified as occurring in both the
selected public and private assistance rolls, the set was then merged
back with the IM and/or Food Stamp monthly extract files. This merging
or rejoining of the matched set by SSN with the IM and/or Food Stamp
files enabled the examination of additional information concerning
those receiving both public and private assistance (that is, joint
recipients). For example, the number of cases or households represented
in the matched set of SSNs could be determined by totaling the number
of unique case numbers corresponding to the matched and rejoined set
of joint recipients. In addition, by separately totaling the amount of
public and private cash assistance received over the selected period
by a household, it was possible to determine the amount by which a
public assistance recipient household supplemented its assistance
income with private assistance grants.
By totaling the number of joint recipient households by their
zip codes, maps were generated to geographically display the distribution
of the joint recipient households. These maps were also overlaid with
the location of MAAC agencies and graduated scales showing the number
of services distributed to joint recipients per agency.
Preliminary Findings
----------------------------------------------------------------------------
The first step in the analyses was to determine the amount of joint
usage of public and private assistance. The amount of joint assistance
depends on the length of time and types of public program coverage. For
this study, joint assistance was defined as receipt of both types of
assistance within three-month period, specifically from August 1997
through October 1997. For convenience, this period will be referred in
this report as the third quarter of 1997. Moreover, joint assistance
was estimated by successively broadening public programmatic coverage from:
TANF cash cases (TANF)
TANF cases or Food Stamp-only cases
Public cash assistance from all sources or Food Stamp-only cases
The results of the joint usage analysis on the linked records for
the three-county area are presented in Tables 1 and 2 and Maps A through E.
The following bullets highlight important findings from this analysis.
The number of recipients served by MAAC nearly equals the number
of households receiving TANF (see Table 1).(2)
In the third quarter of 1997, there were 13,430 TANF households and
13,153 MAAC recipients.
Within a calendar quarter, joint users represent a slightly larger
share of the MAAC caseload than the TANF caseload.
There were 2,847 persons in 2,565 households that received assistance
from both public and private sources during the third quarter of 1997.(3)
Joint recipients accounted for 19.1 percent of the TANF caseload and
21.6 percent of the MAAC caseload.
Food Stamp-only recipients were as likely to use MAAC assistance
as TANF households during the third quarter of 1997.
Adding Food Stamp-only recipients to the TANF caseload increased the
number of public assistance recipiency households by 17,985, more than
doubling the size of the public assistance caseload.
Table 1. Joint Public and Private Assistance Recipients
Cash Assistance or Food Stamps and MAAC
(Jackson, Clay, and Platte Counties)
Aug - Oct 97
Recipient Group Unit 3 months
TANF Cash Assistance and MAAC
TANF Cash Assistance Households Households 13,430
Households also Receiving MAAC Assistance Households 2,565
Co-Recipient Households Percent 19.1%
MAAC Assistance Recipients Individuals 13,153
Recipients also Receiving TANF Cash Individuals 2,847
Assistance
Co-Recipients Percent 21.6%
TANF Cash Assistance or Food Stamps and MAAC
TANF Cash Assistance/Food Stamps Households 31,415
Households
Households also Receiving MAAC Assistance Households 6,113
Co-Recipient Households Percent 19.5%
MAAC Assistance Recipients Individuals 13,153
Recipients also Receiving TANF Cash Individuals 6,662
Assistance/Food Stamps
Co-Recipients Percent 50.7%
Cash Assistance (all sources) or Food Stamps and MAAC
Cash Assistance/Food Stamps Households Households 44,385
Households also Receiving MAAC Assistance Households 6,562
Co-Recipient Households Percent 14.8%
MAAC Assistance Recipients Individuals 13,153
Recipients also Receiving Cash Assistance/Food Individuals 7,057
Stamps
Co-Recipients Percent 53.7%
Notes: Numbers represent counts of unique households (or individuals)
of the identified recipient group type which received at least one
assistance grant at ANY point during the three months.
Cash assistance from all sources includes TANF, General Relief,
Medical Assistance, Aid to the Blind, Old Age Assistance, and OASDI
grants to individuals. SSI-only recipients not included.
Households receiving multiple cash grants and/or food stamps are
counted once.
Sources: Data furnished by the Missouri Department of Social Services
and the Mid America Assistance Coalition. Data compiled by Midwest
Research Institute.
By subtracting the number of TANF cases from the sum of the TANF
and Food Stamp-only cases, we find that there were 3,548 Food
Stamp-only MAAC households, or 19.7 percent of the Food Stamp-only
caseload. This was a joint recipiency rate comparable to that for
TANF households.
The Food Stamp-only MAAC users accounted for 29 percent of the
MAAC caseload.
Using the TANF-plus-Food-Stamp-only definition of public
assistance, joint users of MAAC services accounted for one-fifth of
this public assistance caseload, but accounted for fully half of
the MAAC caseload.
The rate of joint recipiency with MAAC agencies among the TANF-plus-Food-
Stamp-only public assistance population was 19.5 percent. This group of
public assistance recipients accounted for an even higher share of MAAC
recipients--50.7 percent.
TANF and Food Stamp-only recipients are more likely to be joint
MAAC users than recipients of other public cash assistance programs.
Including cash assistance cases from non-TANF sources adds another
12,970 households to the definition of public assistance, for a total
public caseload of 44,385 in the third quarter of 1997.
By subtracting the number of TANF-plus-Food-Stamp-only cases from
the total of all public cash and Food Stamp-only cases, we find that
the "other cash assistance" recipients included just 449 joint
household users of MAAC assistance. The "other cash assistance"
population had a joint MAAC recipiency rate of just 3.5 percent and
accounted for 3.0 percent of the MAAC caseload.(4)
Altogether, there were 7,057 persons in 6,562 households who
received both public and private assistance during the third quarter
of 1997. Due to the low level of joint recipiency among "other cash
assistance" recipients, the overall rate of joint recipiency was
14.8 percent of the public caseload. However, all public assistance
recipients accounted for 53.7 percent of the MAAC caseload.
Table 2 presents data on the income assistance packages of joint
public and private recipients.
Table 2. Joint Public and Private Assistance Grants
TANF Cash Assistance and MAAC
(Jackson, Clay, and Platte County Households,
August through October 1997)
Total Households Size of Household
Grant Source and Amount Number Percent 1-2 3 4+
Persons Persons Persons
TANF Cash Assistance
$300 or less 287 11.2 143 99 45
$301 - $500 464 18.1 284 85 95
$501 - $700 239 9.3 74 105 60
$701 - $900 1,044 40.7 292 528 224
$900 or more 531 20.7 0 0 531
Total 2,565 100.0 793 817 955
Median Value $702 $468 $702 $976
TANF Cash Assistance and MAAC Combined
$300 or less 102 4.0 68 18 16
$301 - $500 357 13.9 182 110 65
$501 - $700 333 13.0 182 96 55
$701 - $900 629 24.5 305 234 90
$900 or more 1,144 44.6 56 359 729
Total 2,565 100.0 793 817 955
Median Value $841 $645 $857 $1,090
Increase in Median $139 $177 $155 $114
Value Resulting
from Receipt of
MAAC Assistance
Note: Households receiving multiple cash grants are counted as one
household; however, all cash grant assistance received over the
three-month period is summed per household.
Sources: Data furnished by the Missouri Department of Social Services
and the Mid America Assistance Coalition. Data compiled by Midwest
Research Institute.
MAAC services add $139 to joint recipient TANF households, or
17 percent of their total income assistance package (see Table 2).
During the three-month period of August through October of 1997,
the median value of TANF cash assistance for joint recipient households
was $702. The median ranged from $468 for one-to-two person households
to $976 for TANF households containing four or more individuals.
Across all household sizes, the median value of MAAC assistance
to joint recipient households was $139. The median value of MAAC
assistance ranged from $177 for one-to-two person households to $114
for TANF households of four or more individuals.
MAAC services comprised larger shares of the total income
assistance packages for one-or-two person households than for larger
households. It comprised 27 percent of the total income assistance
for one-to-two person households compared to 10 percent for households
with four or more people.
Maps A through E show the distribution of joint users in the
three-county area, the location and volume of service at MAAC agencies,
and the location of Comprehensive Neighborhood Service sites.
Joint-user households are found in each of the three counties,
but they are concentrated in the urban core of Jackson County.
Map A presents the distribution of joint public and private
assistance households by zip code. Joint users are found in each of
the three counties. They are concentrated, however, in Jackson County,
particularly its urban core, and in the Clay County zip code that
contains Excelsior Springs and surrounding area.
Map B presents a more detailed depiction of joint recipient
households in the urban core of Jackson County. Zip codes having
the highest number of joint users include 64106, 64127, and 64130.
Service sites are generally located in the zip codes having the
highest numbers of joint users.
Map C shows the location of MAAC agencies (white stars) and
Comprehensive Neighborhood Service (CNS) sites (dark stars), as well
as the distribution of joint recipient households. Service sites are
generally located in the zip codes having the highest numbers of joint
users. Some MAAC agencies have a requirement that clients live in their
neighborhoods. Therefore, the location of the service sites probably
facilitates higher levels of joint usage in these areas.
Map D shows detail on the location of MAAC agencies and CNS
service sites in the urban core of Jackson County.
Map E shows, for the urban core of Jackson County, the volume
of services to joint recipients by MAAC agencies.(5)
Next Steps
----------------------------------------------------------------------------
The final report will include the results of two more types of analyses.
First, we will replicate the concurrent findings for an earlier period,
either prior to or soon after the implementation of the first welfare reform
activities in Missouri. Second, we will address the use of private assistance
as public recipients leave the public caseloads, using this unique new data set.
We will examine their patterns of joint recipiency during and after their stay
on public assistance. With this analysis, we will be able to assess the
association between the imposition of sanctions and the usage of MAAC
services and describe the pattern of MAAC usage after cases exit the public
assistance rolls. If feasible, we will forecast the usage of private
assistance as welfare reform becomes more institutionalized. This information
will support community development of private assistance services.
Appendix
----------------------------------------------------------------------------
Procedures for Protecting Confidentiality
In order to find out about the use of private assistance by individuals
who leave the public assistance caseload, the LINC welfare-to-work research
project will match records from the MAAC database with records from the
state's AFDC files. To ensure an accurate identification of the same
individual in both databases, cases will be matched on SSN, date of birth,
and either name or address.
Risk
Under current and past regulations, private in-kind and third-party
assistance (e.g., utility payments) are not counted as income for the
purposes of determining eligibility or benefit level for AFDC/TANF or
Food Stamps. This study will examine private and public assistance
recipiency during the period 1994 to mid-1997.
There is a concern that such assistance could in the future be counted
as income under TANF regulations and that matching records could jeopardize
individuals' eligibility for benefits or change the amount of assistance
due a family. The discovery that some individuals has received private
assistance that had not been reported to the state would result in those
individuals having to repay the excess benefit to the state.
Confidentiality Protections
MRI has a full appreciation of the confidential nature of the databases
to be used in this project. We have over 40 years of experience in dealing
with confidential data and confidential clients. MRI has established
protocols for working with confidential data, and staff are trained in
maintaining data security and confidentiality.
MRI's Security Officer is also the designated Automated Information
System Security Officer. Current security systems cover government classified
and non-government confidential information. Protection of proprietary
information and release of information are covered in the MRI Standards of
Conduct pamphlet and corporate operating policies and procedures.
Databases that contain classified or confidential information require
safeguards to assure nondisclosure of information for individual records
in the system. The security of electronic information is protected through
controlling access to authorized users, masking identities, and password-
protecting files for transfer. MRI has in place measures to foil attempted
internal and external breaches of security. Virus-scanning and system backup
procedures are designed into the local area network (LAN), workstation, and
bulletin board system operations. Key to MRI's Internet implementation is
a firewall system installed concurrently with the domain server equipment
that bars access to MRI databases by external entities.
Special security arrangements for maintaining the confidentiality of
the linked MAAC/AFDC database include:
After the matching procedure is complete and the linked research
database is constructed, MRI will assign new (research project)
identifiers to the matched records. These identifiers will in no way
reference the individual's SSN, AFDC case number, name, or address.
Once the new database is constructed, MRI will destroy our copies of
the original databases. MAAC and LINC officials may be present as the
original records are destroyed.
Access will be restricted to authorized users through a project
password.
No identifying information will be published. Only aggregate
statistics will be released in the project report.
References
----------------------------------------------------------------------------
(1) Only currently active cases were included in the Income Maintenance and
Food Stamp monthly extracts. Cases which received assistance and were closed
within the same calendar month did mot appear on that month's extract.
(2) The average TANF household contain ~ 2.8.
(3) For convenience, the text refers to TANF cases as households. This draws
more clearly the distinction between MAAC cases, which are individuals, and
TANF cases, which are households.
(4) Approximately 70 percent of the "other cash assistance" population are
recipients of Old Age Assistance and Disability payments.
(5) In this grey-scale map, it is difficult to see the dark stars (CNS sites)
in the zip codes with the highest numbers of joint recipients. There is one
CNS site in zip code 64130 and one on the boundary of zip code 64106.
............................................................................
To print: Select File --> Print from your browser's menu.
Return to data section