The Impact of Matched Savings Programs in Korea: Results from the Hope Growing Account Program

Background: Asset-based social policies and programs are in a period of emergence and testing throughout Asia (Sherraden et al., 2014). The Korea government has recognized the limitations of the income-based policy, namely as people struggle to exit poverty. As a new initiative, the Korean government introduced the 'Hope Growing Account (HOPE) program' in 2010 which combines a monthly earned income grant and matched funds for saving (match rate 1:1) to encourage the working poor to build an asset over three years. The program grew quickly: the number of participants increased from 10,000 in 2010 to 32,000 households in 2013. Asset building policies normally aim to promote saving and asset accumulation. Additionally, many stakeholders are also interested in understanding the program’s impact on household income. This study is guided by the following research question: What was the effect of HOPE program participation on household income?

Methods: We used a two group quasi-experimental design. Longitudinal data  (2010 to 2012) from the Hope Growing Account Panel Study and the Korean Welfare Panel Study were analyzed. Each year’s sample includes over 1,308 households. Our analysis implemented a difference-in-differences estimation technique to compare the treatment group and comparison group household monthly earned income at baseline and two years later. To estimate the treatment effect, we used a Propensity Score Analysis method (Guo & Fraser, 2010) to match Hope participations to similar non-participants. For matching, we chose socio-demographic characteristics of household and household heads as conditioning variables: family composition, region, baseline earned income, gender, marital status, age, education, health status, and employment.

Results: Findings revealed that between 2010 and 2012, the treatment group’s mean household monthly earned income increased by US $54.08, whereas that for the non-treatment group decreased by $99.86 (i.e., minus $99.86). The weighted adjusted difference-in-differences estimation showed a large difference between the two groups, $167.75, meaning that the average change of the households’ monthly earned income for the treatment group was $167.75 higher (or better) than that for the non-treatment group. The bootstrap result showed that the difference was statistically significant with 95% confidence interval.

Conclusions and Implications: Our findings indicate that the Hope Growing Account positively impacts participants' earned income. This impact on household income contributes to a knowledge gap in the literature about effects of participation in matched savings programs. We describe how future research needs to clarify the causal mechanisms contributing to the observed increase in income. Future research is also needed to better understand within-group variation in the program impacts. For interventions, we suggest complementary case management to target other non-income aspects of social development. Further, there is an opportunity to combine elements of the HOPE program with active labour market policies, such as the Employment Success Package program, and the Korean equivalent of the Earned Income Tax Credit.