Primary Outcomes (explanation)
Our primary outcome variables of interest were the following. First, we were interested in measuring the impact of our educational intervention on whether or not participants open a myRA. Our measure of whether or not participants open a myRA wass self-reported, which is a limitation of our study, and we discussed this later in the paper. Because we helped participants to open myRA on the computer at the end of the workshop, we were able to know with more certainty whether or not the participants opened the account at the workshop.Footnote11
Second, we were interested in determining the impact of our intervention on knowledge related to financial planning for retirement. During the design stage of our study, we found that many existing retirement knowledge and planning scales were more applicable to individuals with a middle to high socio-economic status. For example, the scale most recently created by Hopkins and Littell (2016) featured items that require a high-level of education and literacy for individuals to fully comprehend the questions. Hetling et al. (2016) developed a financial literacy scale that included a subscale on investing and long-term planning knowledge. We found that this scale was not useful for our study when we first piloted our survey among our target population because many items in the scale were too complex and difficult to comprehend.
We created a simple indicator that measured self-reported knowledge related to retirement planning. Our indicator was a combination of questions from the HRS related to knowledge about retirement planning and saving, similar to those used by Lusardi and Mitchell (2011a), and the indicator created by Carr et al. (2015). Table A1 in the Online Supplemental Material shows the components of the RKI, which consisted of 10 yes/no questions. Therefore, our indicator has a 0–10 value. We also considered a modified retirement knowledge indicator (M-RKI) that takes into consideration the fact that question 8 asks if a participant has attended a workshop about retirement planning. In the follow-up workshop we clarified and asked whether participants attended other workshops besides the one associated with this program (see Table A1 in the Online Supplemental Material for specific question modification). However, it might be the case that participants counted our initial workshop in the scale and answered “yes” to this question. Thus, we estimated our indicator without this question, so that our modified retirement knowledge indicator has a 0–9 value.
Third, we also evaluated whether our intervention had any effect on participants’ financial management behavior. We used the indicator constructed by Dew and Xiao (2011). This scale is composed of the following subscales: (1) saving and investment, (2) cash management, and (3) credit management.Footnote12 This indicator was not ideal in our sample as many participants answered that four of the questions did not apply to them (see Table A2 in the Online Supplemental Material for questions used in the scale and subscales). The questions where participants answered that it did not apply to them were those related to the credit management subscale because many participants did not have credit cards or have outstanding loans. We also had several participants who answered that saving from each paycheck was not applicable since they were not working in the last six months. In our Online Supplemental Material we included our English and Spanish Baseline Surveys.