Psychological determinants of household saving behaviour

Abstract

This dissertation reports the results of a study that examined the impact of psychological variables on household saving and borrowing behaviour. Understanding saving behaviour is important for policy makers and financial institutions, but a comprehensive explanatory model that can explain individual differences in saving does not yet exist. The aim of this research has been to contribute towards the construction of such a model. The study was designed to answer four research questions. The first research question examined psychological explanations for saving in the existing economic and, economic psychological literature on household saving in order to determine the strength of their empirical support. Though previous reviews have identified several theories, based on psychological ideas, they have not systematically evaluated the theories empirically. This is important because it establishes which theories should be subject to further investigations and which should be ignored. The reviews show that many of the psychological theories have been tested only once. In addition, many studies suffer from serious weaknesses with respect to the measurement of variables. It is therefore appropriate to say that psychological research on household saving is in its infancy and that the fundamental question - which role psychological variables play in household saving - is left unanswered. Consequently, the second research question concerned the contribution of psychological variables in explaining variation in household saving. This question was answered by conducting an empirical study using Dutch data collected for the CentER Saving Survey. The results of the empirical study demonstrate that psychological variables are well worth considering when trying to describe, explain, and predict saving and borrowing behaviour. In particular, time horizon an attitudes towards debt are found to be the most significantly associated with saving and borrow behaviour.

The third research question addressed the notion that psychological variables i influence different types of saving. If this is the case, these relationships will not necessarily emerge when total saving is analysed in isolation. Several different measures of saving behaviour were therefore used. More specifically, a distinction was made between discretionary and contractual saving and between flow (saving during a year) and stock (wealth at a particular moment in time) measures of saving. Discretionary saving includes saving that is rely to be a result of discrete saving decisions in the accounting period. Contractual saving, defined here, includes repayment of debt and mortgages, and results from previous borrowing decisions. The data in this thesis suggest that it is useful to distinguish between different types of saving and wealth when investigating the effects of psychological variables.

The fourth and final research question concerned the possibility that psychological variables may have more impact on saving behaviour among high-income households in comparison with low-income households. As suggested by Katona (1975), the substantial increase and spread of discretionary income among families in Western economies may increase the importance of variables other than income when explaining saving and spending decisions. This discretionary income means that there is income left to spend after necessities and habitual expenses are paid for, so that a choice between spending and saving can be made. It is reasonable that the increase in discretionary income alters the relative influence of economic, situational and psychological variables on saving in favour of psychological variables. In order to test this assertion, estimates based on data from three different discretionary income groups were compared. The results did not support the hypothesis that the psychological variables were more important predictors for the saving behaviour of the richer than the poorer part of the sample. Rather, the results suggest that psychological variables are im t for explaining saving behaviour at all income levels. The thesis close with a discussion of the strengths and weaknesses of the study, as well as the implications of the results.

This dissertation reports the results of a study that examined the impact of psychological variables on household saving and borrowing behaviour. Understanding saving behaviour is important for policy makers and financial institutions, but a comprehensive explanatory model that can explain individual differences in saving does not yet exist. The aim of this research has been to contribute towards the construction of such a model. The study was designed to answer four research questions. The first research question examined psychological explanations for saving in the existing economic and, economic psychological literature on household saving in order to determine the strength of their empirical support. Though previous reviews have identified several theories, based on psychological ideas, they have not systematically evaluated the theories empirically. This is important because it establishes which theories should be subject to further investigations and which should be ignored. The reviews show that many of the psychological theories have been tested only once. In addition, many studies suffer from serious weaknesses with respect to the measurement of variables. It is therefore appropriate to say that psychological research on household saving is in its infancy and that the fundamental question - which role psychological variables play in household saving - is left unanswered. Consequently, the second research question concerned the contribution of psychological variables in explaining variation in household saving. This question was answered by conducting an empirical study using Dutch data collected for the CentER Saving Survey. The results of the empirical study demonstrate that psychological variables are well worth considering when trying to describe, explain, and predict saving and borrowing behaviour. In particular, time horizon an attitudes towards debt are found to be the most significantly associated with saving and borrow behaviour.

The third research question addressed the notion that psychological variables i influence different types of saving. If this is the case, these relationships will not necessarily emerge when total saving is analysed in isolation. Several different measures of saving behaviour were therefore used. More specifically, a distinction was made between discretionary and contractual saving and between flow (saving during a year) and stock (wealth at a particular moment in time) measures of saving. Discretionary saving includes saving that is rely to be a result of discrete saving decisions in the accounting period. Contractual saving, defined here, includes repayment of debt and mortgages, and results from previous borrowing decisions. The data in this thesis suggest that it is useful to distinguish between different types of saving and wealth when investigating the effects of psychological variables.

The fourth and final research question concerned the possibility that psychological variables may have more impact on saving behaviour among high-income households in comparison with low-income households. As suggested by Katona (1975), the substantial increase and spread of discretionary income among families in Western economies may increase the importance of variables other than income when explaining saving and spending decisions. This discretionary income means that there is income left to spend after necessities and habitual expenses are paid for, so that a choice between spending and saving can be made. It is reasonable that the increase in discretionary income alters the relative influence of economic, situational and psychological variables on saving in favour of psychological variables. In order to test this assertion, estimates based on data from three different discretionary income groups were compared. The results did not support the hypothesis that the psychological variables were more important predictors for the saving behaviour of the richer than the poorer part of the sample. Rather, the results suggest that psychological variables are im t for explaining saving behaviour at all income levels. The thesis close with a discussion of the strengths and weaknesses of the study, as well as the implications of the results.