Title: Cash
1Cash Counseling Demonstration Evaluation
Randall S. Brown, PhD Mathematica Policy
Research, Inc. December 14, 2005
2Original Cash Counseling Demonstration Overview
- Demonstration States
- Arkansas, Florida, New Jersey
- Study Populations
- Adults with disabilities (Ages 18-64)
- Elders (Ages 65)
- Florida only Children with developmental
disabilities - Feeder Programs
- Arkansas and New Jersey Medicaid personal care
option programs - Florida Medicaid 1915c Home and Community-Based
long-term care waiver programs
3Study Design and Methods
- Randomly assigned applicants (AR, FL, NJ)
- 1,700 - 2,000 adults per state, 1,000 children
(FL) - Data
- Consumer survey at 9 months
- Primary unpaid caregiver survey at 10 months
- Survey of paid workers at 10 months
- Medicaid claims for two years
4Receiving Paid Assistance at 9 Months
Non-Elderly Adults
Percent
Elderly Adults
Children
T C NJ
T C FL
T C NJ
T C FL
T C AR
T C FL
T C AR
, Significantly different from control group
at .05, .01 level, respectively.
5Very Satisfied with Overall Care Arrangements
Percent
Non-Elderly Adults
Children
Elderly Adults
T C NJ
T C NJ
T C FL
T C AR
T C FL
T C AR
T C FL
, Significantly different from control group
at .05, .01 level, respectively.
6Had an Unmet Need for Help with Personal Care
Percent
Non-Elderly Adults
Children
Elderly Adults
T C NJ
T C FL
T C NJ
T C FL
T C AR
T C FL
T C AR
, Significantly different from control group
at .05, .01 level, respectively.
7Contractures Developed or Worsened
Percent
Non-Elderly Adults
Children
Elderly Adults
T C NJ
T C FL
T C NJ
T C FL
T C AR
T C FL
T C AR
, Significantly different from control group
at .05, .01 level, respectively.
8Very Satisfied with Way Spending Life These Days
Percent
Non-Elderly Adults
Children
Elderly Adults
T C NJ
T C FL
T C NJ
T C FL
T C AR
T C FL
T C AR
, Significantly different from control group
at .05, .01 level, respectively.
9Informal Caregivers Very Satisfied with Overall
Care
Adults
Children
Percent
T
T
T
C
C
C
T
C
AR
FL
FL
NJ
, Significantly different from control group
at .05, .01 level, respectively.
10Informal Caregivers Experienced Less Emotional
Strain
Adults
Children
Percent
C
T
T
T
C
C
T
C
AR
FL
FL
NJ
, Significantly different from control group
at .05, .01 level, respectively.
11Informal Caregivers Experienced Less Physical
Strain
Adults
Children
Percent
T
T
T
C
C
C
T
C
AR
FL
FL
NJ
,, Significantly different from control
group at .10 (), .05 (), or .01 () level.
12Informal Caregivers Experienced Less Financial
Strain
Adults
Children
Percent
C
T
T
T
C
C
T
C
AR
FL
FL
NJ
, Significantly different from control group
at .05, .01 level, respectively.
13Relationship of Consumer toDirectly Hired Worker
Percent
AR Adults
FL Adults
NJ Adults
FL Children
14Hours of Care Provided by Directly Hired Workers
- Paid Hours per Week
- 12 (Arkansas) to 20 (Florida, New Jersey)
- Unpaid Hours per Week
- 59 to 74 provided some unpaid hours
- 25 provided more than 41 unpaid hours
15Working Conditions
16Training and Preparedness of Directly Hired
Workers
- Half trained in personal care or routine health
care - On the other hand
- Felt well-informed about care recipients
condition (90) - Felt fully prepared for job (gt90)
- Injuries and physical strain similar to agency
workers providing equal hours of care
17Worker Well-Being
- Directly Hired Workers More Likely to
- Experience emotional strain
- Want more respect from consumers family
18Worker Well-Being (contd)
Percent
Feels Emotional Strain
Wants More Respect
19Effects on Medicaid PCS/HCBS ExpendituresYear 1
- Significantly Higher for Treatment Group in Each
State - In AR and NJ, Mainly Because Control Group
Received Substantially Less Care Than Authorized - In FL, Mainly Because Children and Adults With
Developmental Disabilities Got Larger Benefit
Increases After Assigned to Treatment Group
20Effects on Medicaid PCS/HCBS ExpendituresYear 1
Significantly different from zero at .01 level.
21Effects on non-PCS Medicaid Expenditures
- Other Medicaid Costs Moderately Lower For
Treatment Group in Each Age Group in All Three
States - The Best Example
- In AR , Compared to Control Group, Treatment
Group Had 40 Fewer Admissions to Nursing
Facilities in Second Year
22Effects on non-PCS Medicaid ExpendituresYear 1
Significantly different from zero at .05 level.
23Effect on Total Medicaid Costs
- In AR, No Significant Difference by End of Year 2
- Reductions in NF and other Waiver Costs Off-Set
Increase in Personal Care Costs - In NJ and FL, Costs Up 8-12, But States Learned
How to Control Costs - Higher Costs in AR and NJ Due to Failure of
Traditional System
24Effect on Total Medicaid CostsYear 1 vs. Year 2
Significantly different from zero at .01 level.
25Policy Implications
- Can increase access to care
- Greatly improves quality of life (all ages)
- Caregivers also benefit greatly
- States may be concerned about costs
- But have learned how to control them
26Lessons on Controlling Costs
- Consider limiting program to current recipients
- Use standardized assessment tool and independent
staff to set allowance - Adjust allowance amount to reflect expected use
- Use incentive-based contracts for counseling
services - Set clear and fair recoupment policy for unspent
allowances - Monitor reassessments and costs