Financial well-being in an urban area: an application of multiple imputation

Abstract

This article estimates a model of self-reported financial well-being (FWB) using primary data collected for a Southwestern U.S. city. Missing data are estimated using multiple imputation. Model estimates show how FWB depends on home ownership, the number of children, health insurance, age, and income. Multiple imputation results differ somewhat from complete case results.

Acknowledgements

The author thanks the anonymous referee and Joachim Zietz for many helpful comments. Data collection for this study was supported by a contract with Community Council of Central Oklahoma, Inc. An earlier version of this article was presented at the Academy of Economics and Finance annual meeting in Myrtle Beach, South Carolina, in February 2005.

Notes

¹DiTella and MacCulloch (2006) and Kahneman and Krueger (2006) provide overviews of happiness and SWB

²Penn (2007) provides an example of using MI to estimate missing values in the General Social Survey.

³Raghunathan (2004) is a recent example in the public health literature. Schafer and Graham (2002) provide an excellent discussion of the principles and applications of MI.

⁴An earlier example in this journal of a study of the use of MI in this journal may be found in Keng et al. (2002).

⁵IVEware software and documentation are available at http://www.isr.umich.edu/src/smp/ive/.

⁶More discussion of the study and general descriptive statistics may be found in Penn (2005).

⁷Since college attainment is included as an explanatory variable, the sample is limited to ages 25 and older.