Saving Performance in an IDA for Participants with Disabilities: Do Institutional Characteristics Matter?

background and purpose: Asset development policies have been promoted as a means to create a more inclusive “ownership society”. During the past few years asset building scholarship has begun to focus specifically on marginalized groups including persons with disabilities. Guided by institutional theory, this study asks: 1) If disability status impacts savings in an IDA (Average Monthly Net Deposit or AMND); 2) If yes, do (a) individual and household characteristics; along with (b) institutional characteristics, influence the impact of disability on AMND.

Methods: These questions are answered using a sample of Individual Development Account (IDA) program participants (N=376) from the American Dream Demonstration's (ADD) IDA random controlled trial in Tulsa, OK. Participant data were collected as part of the ADD study from 1997-2003. MIS IDA program administrative data provide institutional measures including IDA design and saving outcomes. This is the first study of an ADD, IDA program focusing on participants with a disability. Key study variables include disability status, individual and household characteristics, and program (institutional) factors – participation in program activities, staff contact, asset specific and general financial education.

Univariate and bivariate procedures were used to assess differences between participants with and without disability by savings outcomes, individual and household characteristics, as well as program (institutional) factors. A three step sequential regression procedure was employed that treated disability, individual and household characteristics, in addition to program factors as distinct variable blocks to assess their cumulative relationship to AMND.

Results: Results indicated that program participants with disabilities were significantly more likely to be female, older, and unemployed. There were no differences in household characteristics by disability status. However, program participants with disabilities had significantly lower AMND amounts than participants without disabilities ($20.72 vs. $26.07). Significantly smaller percentages of the disability sub-sample participated in asset specific financial education (34% vs. 45%). In assessing relationship to AMND, Model I indicated that participants with disabilities saved about $5.50 less than participants without disabilities (about $400 per year). In Model II, several individual and household characteristics, including age, education, race, and household size, were significant. After controlling for these differences, disability status remained significant with AMND difference increasing to $6.60. Model III added program characteristics (participation in program activities, staff contact, and general financial education) and showed their joint effects to be significant. Variables significant in Model II, including disability status remained so in Model III. The inclusion of institutional variables in Model III buffered about 1% of the variance in AMND that is explained by disability status.

Conclusions and implications: Conclusions drawn from this study are that disability status, in addition to a number of individual and program characteristics, is associated with saving in an IDA. Ultimately, lower savings for IDA participants with disabilities reduced total amount of match funds received. IDA programs aimed at mediating poverty among people with disabilities should consider the unique role disability status (the experience of living with a disability and receipt of disability-related benefits) may play in the successful participation of people with disabilities in IDA programs.