Abstract
The authors use data from a finance-related theme park to explore how financial education changes investment, financing, and consumer behavior. Students were assigned fictitious life situations and asked to create household budgets. Some students received a 19-hour financial literacy curriculum before going to the park, and some did not. After controlling for demographic variables, the authors show that the treatment effects of the financial literacy program are strong. Students were more frugal, delayed gratification, paid off debt faster, and relied less on credit financing after training. Students who attended training showed greater uptake of decision support that was offered in the park, which indicates that decision support and financial literacy training are complements, not substitutes.
Keywords :
Acknowledgments
The authors thank Viral Acharya, Doug Bernheim, Annamaria Lusardi, John Lynch, Adair Morse, Bill Walstad, and seminar participants at Vanderbilt University, Georgetown University, Duke University, the 2012 American Economic Association meetings, and the 2010 Fall NBER Corporate Finance meeting for helpful comments and feedback. The authors also thank and acknowledge colleagues at Junior Achievement of Southern California for helpful discussions and guidance throughout the project. This work was conducted in compliance with IRB human subject guidelines and was approved by the IRBs at Duke and UCLA. Dan Blanchette provided excellent research assistance.
Notes
1. One exception is Bernheim and Garrett (Citation2003), who show that financial education leads to higher aggregate savings; however, behavior at the individual level is not measured in their study.
2. A full description of the curriculum is available in Carlin and Robinson (Citation2009), and at http://www.jasocal.org.
3. Examples are taken from the JA Finance Park Student Workbook, which is available upon request.
4. Additional details can be found online at http://www.nber.org/papers/w16271.
5. Tables are available upon request.
6. Confidentiality requirements prevent us from disclosing the name of the school. However, it is in the South Los Angeles area, and is approximately 52 percent black and 47 percent Hispanic in student makeup. It has an API score of 1, placing it in the lowest decile, and has around two percent of students at or above California state-level math proficiency.
7. These results can be found in greater detail online at http://www.nber.org/papers/W16271.
8. This question is also explored in greater detail in Carlin and Robinson (Citation2012).
9. Difficulty in extrapolating beyond the scenarios used to illustrate educational principles is discussed in the educational psychology literature. See, for example, Thompson, Gentner, and Lowenstein (Citation2000) or Gentner, Lowenstein, and Thompson (Citation2003).