Financial literacy has been recognised as a vital life skill, but there is little evidence of the factors behind the differences in managing personal finance. Socio‐economic factors and the provision of financial education do explain the variance in financial literacy in some countries, but not in all. In the PISA 2012 financial literacy test, Estonian students ranked very highly in international comparison; although only a few had received financial education at school. Compared with other countries, socio‐economic factors explained the smallest proportion of variance in the test score. There was, however, a significant difference between the mean financial literacy scores of Estonian‐ and Russian‐language communities. The aim of the article is to analyse the factors behind the differences in financial literacy when financial education is not provided. It also offers insight into how students in a similar education system in two different cultural and language frameworks achieve different financial literacy scores. Moreover, the results demonstrate how indicators, such as family background can work through different channels as opposed to the usual parental education or occupation based socio‐economic indicators. The latter implies that unexplained factors remain, such as cultural, developmental and societal indicators, which most researchers pay little attention to when explaining efficient policies for improving financial literacy. Multivariate regression models show that the level of financial literacy in Estonia is correlated with gender, language of the school, the number of books at home, mathematics and reading scores. The Blinder–Oaxaca decomposition explains less than half of the gap between the two communities. The only variable significantly explaining the gap is the number of books at home. Books can be interpreted as a symbol of social status, evidence of cultural background or source of influence for broader picture and better problem solving skills.

No abstract

In recent years, the degree of choice in education systems has increased in most countries. Still, the variation of choice policies across countries is substantial. The authors ask under what combinations of conditions (i.e. institutional features of education systems) choice policy succeeds in balancing educational efficiency and equity. Using the fuzzy-set qualitative comparative analysis, they investigate the impact of seven institutional conditions in 20 European countries. Those seven conditions are identified in school choice literature as relevant in explaining variations in educational efficiency and equity. The analysis shows that there are multiple causal paths to good policy outcome. The main contribution of this article is to show that 'choice' is an INUS condition (i.e. an insufficient but necessary part of an unnecessary but sufficient combination of conditions) and that 'no tracking' is a necessary condition for educational efficiency and equity. In addition, the authors show that 'good management' and 'competition' of schools contribute to good educational outcomes only in choicetolerant countries.

We examine Kindergarten allocation practices in an Estonian municipality, Harku. Based on our recommendations, the allocation mechanism in Harku was redesigned in 2016. The new mechanism produces a child-optimal stable matching, with priorities primarily based on siblings and distance. We evaluate seven policy designs based on 2016 admission data in order to understand efficiency and fairness trade-offs. In addition to the descriptive data analysis, we conduct a counter-factual policy comparison and sensitivity analysis using computational experiments with generated preferences. We fix the allocation mechanism to be the child-oriented Deferred-Acceptance algorithm, but we vary how the priorities are created by altering sibling and distance factors. Different lotteries are included for breaking ties. We find that different ways of considering the same priority factors can have a significant aggregate effect on the allocation. Additionally, we survey a dozen special features that can create significant challenges (both theoretical and practical) in redesigning the

Economists and philosophers disagree about the concept of choice used in economics. Some behavioural economists argue that economic models of choice will improve as they become more and more psychologically realistic. Don Ross argues that this argument fails because its hidden assumption – that the economic concept of choice is the same as the psychological counterpart – is false. Ross conjectures that the economic concept of choice concerns a population-scale pattern of behavioural changes in response to incentives. We conduct a survey experiment to test two predictions that Ross’s conjecture generates. The statistical analysis of our data confirms our predictions, although with some qualifications. In interpreting our results, we distinguish two versions of commonsensible realism, strong and weak, and propose the weak one as a plausible explanation of our results. Weak commonsensible realism also produces further testable hypotheses. Some methodological implications of our study are discussed.