Research That Matters (January 17 - 20, 2008)

Purpose: Much of the research examining Sherraden's (1991) asset effects have focused on the experience of asset holding. Assets could also be experienced in terms of the process of asset accumulation or its consumption (Paxton, 2001). However, little empirical work has been done to examine the effects of the process of accumulation. Moreover, most studies do not differentiate the effects of assets on the different socioeconomic classes, nor do they discriminate between the different asset measures. This paper aims to fill this gap in the knowledgebase by examining the relationship between the process of asset accumulation as measured by the change in assets over time, with children's educational outcomes. In addition, we will examine if the effects are moderated by the level of family income. We hypothesize that the increase in assets over time will be associated with children's educational outcomes, in particular, that the different asset measures will have differential effects for the various income groups.

Methods: Data on 1342 children ages 7 to 14 in the year 2000 and on their mothers drawn from the NLSY79 and the NLSY79 Children and Young Adults datasets are used for this study. The outcome measures are children's PIAT standardized scores in math and reading in 2000. The process of asset accumulation is measured by whether there is an increase in net-worth, net-worth less home equity, and in liquid assets, from 1996 to 2000. Income is measured be averaging the total net family income over 1996 to 2000. Other variables include various mothers' socio-demographic data. A series of OLS regressions by the different income quartiles were conducted to examine the relationship between changes in assets and children's educational outcomes.

Results: The findings support our hypothesis that the different asset measures have differential effects for the various income groups. Increases in net worth significantly predict better math scores for children from the 2nd (b=4.04, t=2.19, p=0.03, N=173) and 4th income quartiles (b=4.70, t=2.08, p<0.04, N=171). Increases in net-worth less home equity is also significantly associated with better math outcomes for the 2nd income quartile (b=4.70, t=2.65, p=0.009, N=180) but lower math scores for children from the 3rd income quartile (b=-4.78, t=-2.21, p=0.03, N=165). Similarly, increases in liquid assets significantly predicts better math outcomes for children from the 2nd (b=3.63, t=2.19, p=0.03, N=201) and 4th (b=4.75, t=2.22, p=0.03, N=193) income quartiles, but lower math scores for the 3rd (b=-4.31, t=-2.03, p=0.04, N=181) income quartile. No significant associations are found for reading scores across the different asset measures. Race and mother's education are also found to significant predict children's math scores in the models reported above.

Implications: Our study calls for additional longitudinal research to further explicate the dynamic relationship between the process of asset accumulation across the different asset measures and children's outcomes for the different income groups. Diverse policy approaches would also need to be introduced to maximize the effects of parental assets on children's outcomes across income status.