Financial education and investment attitudes in high schools: evidence from a randomized experiment

Abstract

We experimentally study the effect of financial education on investment attitudes in a large sample of high school students in Italy. Students in the treated classes were taught a course in finance and interviewed before and after the study, while controls were only interviewed. Our principal result is that the difference-in-difference estimates of the effect of the course are not statistically significant. However, the course in finance reduces the virtual demand for cash and increases the level of financial literacy and the propensity to read (and the capacity to understand) economic articles in both treated and control classes compared with pre-treatment baseline levels. A breakdown of the cognitive process, which is statistically significant for the classes treated, suggests that error and ignorance reduction is sizable and that the progress in financial literacy is stronger in subgroups which exhibit lower ex ante knowledge levels.

Keywords:

• financial education
• financial literacy
• demand for money balances
• randomized experiment

JEL Classification:

• I22
• E41
• G10
• G20

Acknowledgements

We thank Gianpaolo Barbetta, Stefano Cima, Iftekhar Hasan, Tullio Jappelli, James Lothian, Chiara Monticone, Mario Padula, Paul Wachtel and all participants at the XIXth Tor Vergata Financial Conference for their useful comments and suggestions. Managerial support from Marco Bracaglia and the Osservatorio sul Credito, institutional support from the Ministry of Education and financial support from Unicredit, Fondazione Cariplo, Fondazione Carige and JP Morgan are gratefully acknowledged.

Notes

1 To provide an example on this point, Lusardi and Mitchell (2008) underline that baby boomers in the United States are approaching retirement with very low levels of financial wealth, thereby creating a potential factor which is likely to increase individual bankruptcies and the cost of social assistance in the future.

2 For a survey and results on returns to schooling in various countries, see, among others, Psacharopoulos and Patrinos (2004) and Card (1999).

3 According to the OECD, ‘Financial education should start at school, for people to be educated as early as possible’ (OECD, 2006). The importance of financial education has also been noted by the European Commission in the White Paper on Financial Service Policy in the period between 2005 and 2007 and the Green Paper on Retail Financial Service in the Single Market in 2007. The importance of financial education in the United States has been emphasized, among others, by the 2006 review ‘Taking Ownership of the Future’. In the United Kingdom, after verifying the poor performances in primary and secondary schools, the Financial Services Authority (FSA) launched a national plan (the National Strategy for Financial Capability), while other projects (Citizens Advice Frontline Training, On Your Own 2 Feet, Money Doctors) focus on 11- to 19-year-old students and adults. In 2006, the Consumer Financial Education? Body (CFEB) was created with the goal of promoting financial education. In France, the ‘Federation Bancaire Francaise’ has organized courses of financial education in French schools, and in 2006 the Institut pour l'Education Financière du Public was created to coordinate state and private initiatives on financial education. In Germany, the ‘SchülerBanking’ project provides financial education to scholars in collaboration with the Jugend und Bildung association, which includes financial experts, members of labour and industrial unions and representatives of the Ministry of Education. Recently, a joint project has been instituted between the Bundesbank and the Ministry of Finance to create a network of institutions which will support financial lectures in schools. In Spain, the ‘Educación y asesoramiento financiero’ project, a private initiative with economic support from the Ministerio de Educación, Política Social y Deporte, provides financial education to families in order to help them to solve their financial problems.

4 The Liceo Classico has historically been considered the most prestigious type of high school in Italy. Its curriculum is mainly in the humanities (Latin, Greek, Italian and philosophy), but also includes mathematics, physics, chemistry and biology. The Istituto Professionale is a technical school in which the curriculum includes accounting and basic economic principles together with Italian, mathematics, and principles of aw.

5 The details of the structure of the course are omitted for reasons of space, but are available upon request.

6 The remaining heterogeneity due to the influence of class-specific financial education teachers or school teachers (and/or the impact of the socioeconomic environment in which the school is located) will be controlled for in the econometric analysis with class dummies, or by clustering standard errors at class level according to the various specifications estimated (see Section V for details).

7 These grades are on a scale of 0–10, with 6 being the pass level.

8 The questionnaire is in an Appendix available upon request.

9 The puzzle remains, however, since the effect persists when we control for income and parental education in the econometric estimates discussed in Section V (see Table 11). Hence the Liceo Classico effect must be related to a teaching quality factor, or to socioeconomic factors which are not captured by parental job and education variables.

10 Opinions on the comparative value of wrong and DK answers are mixed. According to some, DK answers acknowledge ignorance, and are therefore better than giving wrong answers on the presumption that they are right. In order to see whether our findings were affected by small perturbations of the indicator, we devised a robustness check in which we set to missing transitions from right to right answers. The results were substantially unchanged, and are available upon request.

11 The two questions where we do not register progress in the treatment group are those on the exchange rate and the role of the Bank of Italy. More specifically, in the case of exchange rates the majority of students choose ‘the price you pay to buy foreign goods’ instead of ‘the price paid to exchange domestic with foreign currency’, while in the case of the Bank of Italy, they prefer the answer ‘the Bank of Italy decides the monetary policy in Europe’, to ‘the Bank of Italy is in charge of banking supervision in Italy’.

12 The mistakes are on CONSOB (the Stock Exchange Regulatory Authority), banking foundations, the Bank of Italy, current accounts, mortgages, ATMs, risks and exchange rates.

13 As a robustness check we built the index by considering as missing values situations in which the response was right both before and after the treatment. The results were substantially unchanged, and are available upon request.

14 The result obviously depends in part (but not completely) on the higher potential increase in correct answers for groups with lower ex ante financial literacy.

15 The idea that ignorance (DK answer) may be a better status than error seems to be confirmed by the fact that the transition to correctness is higher (and persistence in the same state is lower) in the first than in the second case, even though more than 10% of ignorance moves to error. Consider, however, that the difference between average error reduction and average ignorance reduction across the 27 questions is not statistically significant.

16 We also tested the effect of the course on the share of students who saved money out of funds received monthly from their parents. The increase in the share of those who saved was 1%. Unfortunately, we do not have information on the share saved. Consider, however, that, differently from the virtual investment question, in the case of their monthly allowance, students received only a very small amount of money, which was explicitly intended to satisfy their expenditure needs, and therefore they may have reasonably believed that savings attitudes must be delegated to parents. Furthermore, it is possible that in several cases parents also saved money for their children over and above their monthly allowance.

17 The results are omitted for reasons of space, and are available upon request.