An open research question is how households respond to expected versus unexpected income shocks. The aim is to study household savings behavior in the presence of expected and unexpected bonuses. The authors test the lifecycle/permanent income hypothesis against the behavioral lifecycle theory.
The lifecycle/permanent income hypothesis is the “workhorse model” that is used to explain the allocation of consumption and savings over the lifecycle (Attanasio and Weber, 2010). The prediction of (several versions of) the permanent income hypothesis is that households will save most of a positive income shock. Moreover, households will save the same fraction of both expected and unexpected bonuses. An alternative is put forward by Shefrin and Thaler (1988) with the behavioral lifecycle theory. Distinguishing ingredients are self-control problems, mental accounting and framing. Self-control refers to the struggle between immediate gratification and long-run benefits. Mental accounting refers to the way individuals mentally organize income and spending. Framing is the way income is labeled by the household, and a couple of studies have found effects of mental accounting on savings behavior; see e.g. Card and Ransom (2011). The behavioral lifecycle hypothesis generates different predictions for saving out of bonus income. If a bonus is expected, the prediction is that households will still save a large fraction of it. A major difference between the permanent income hypothesis and the behavioral lifecycle theory is the treatment of unexpected bonuses – ‘windfalls’. The size of the bonus determines whether the household will treat it as current income (small windfalls), or current wealth (larger windfalls). If it is treated as current income, it will be consumed. When treated as current wealth, it will be saved. Both splurging and saving are possible outcomes. Which of the two theories is the most accurate in describing savings behavior is ultimately an empirical matter.
Statistics Netherlands has released a unique administrative dataset with detailed payroll data for all employees in The Netherlands. For each employee up to 72 months of detailed salary information are available. I will exploit within-employee variation of employees’ bonus income, as well as variation between employees, between firms and between industries. Unexpected bonus income – performance bonuses, profit sharing – can be identified in the data, by exploiting the monthly panel structure of the data. Since the data contain firm and industry identifiers it is possible to compare employees within the same firm over time, firms within the same industry over time – to exploit changes in firm specific labor agreements, as well as industries over time. We will link the employees' wage profiles to an administrative dataset with the household asset information of the employee, in order to differentiate the impact of expected and unexpected bonus income on portfolio choice.
The set-up of the remote access facility with Statistics Netherlands has been completed. The intake process with Statistics Netherlands as well as getting all required permissions from the Dean, the Netherlands Statistics Commission took longer than scheduled. Furthermore, there were some technical difficulties with the remote access software. All issues are now solved and the data can be accessed. The data cleaning started in September 2014 and is in progress now. Furthermore, clean identification of income shocks requires additional data collection. As a source of exogenous variation, we collect information from Collective Labor Agreements. We have acquired a national longitudinal database, and we are in the process of cleaning.
|237||Nathanael Vellekoop||Explaining Intra-Monthly Consumption Patterns: The Timing of Income or the Timing of Consumption Commitments?||2018||Household Finance||consumption, consumption commitments, paycheck frequency, liquidity|