| dc.description.abstract | This paper investigates income and consumer behavior among elderly Norwegians. Analyzing consumer behavior at retirement is important for a number of reasons; the demographic situation of many countries, as in Norway, is characterized by a growing elderly population . It is therefore imperative to devote time to study effects of this situation. Studying consumption of the elderly allows the direct estimation of how well Social Security retirement benefits meet one of the main goals of the program: the maintenance of consumption. A reform of the retirement scheme is planned in Norway , and analyzing decision making at retirement can add some new insight as to how well off the elderly in Norway are. In addition, it is of substantial interest to test the predictions of the life-cycle model in order to answer fundamental questions on how households make economic decisions about the future.
The text book version of the life-cycle model predicts a smoothed consumption path over the life cycle, and therefore consumption should be constant over the span of life. It’s assumed on that people will save in periods when their income is high relatively to average, and the opposite when income is relatively low. Consumption should be constant and the behavior of people in the model is based on expectations of future income not present income. People expect a drop in income at retirement with an income higher than consumption before retirement, and income lower than consumption after retirement. People save to order to consume in the future.
The key ideas of the life-cycle hypothesis come from the work of Modigliani and Brumberg (1954) and Friedman (1957). The model is the standard model for the analysis of inter-temporal decision making at the household level. There has however been a great deal of work done in the analysis of possible limitations of this model. Many researchers find the model useful, but have also had difficulties reconciling the model with empirical findings.
Common for studies analyzing spending at retirement, is that a significant decline of individuals/households consumption at retirement age is observed. This decline is associated with a reduction in income and can be interpreted as an indicator for inadequate resources at retirement resulting in reduced consumption. However, it might also be evidence of inadequate financial preparations for retirement. The decline can not be explained by the life-cycle model, since the life-cycle model predicts constant consumption. This phenomenon is called the retirement-consumption puzzle and has been the topic of a number of investigative studies. In the US, Hamermesh (1984) utilized the life-cycle model to examine 500 Caucasian, married couples and concluded that average consumption exceeds 14 percent of the total income in their retirement years. He concludes that saving for retirement is insufficient to sustain the level of consumption.
A number of other studies try to make new assumptions and versions of the life-cycle model in order to attempt to reconcile with the puzzle. Haider and Stephens Jr. (2004) argue that subjective expectations should be utilized instead of rational expectations in the life-cycle framework. They find that retirement expectations are significantly correlated with consumption changes. When using retirement expectations as an instrument for retirement, they find that consumption falls by 7 to 11 percent for workers who retire as expected. Hurd and Rohwedder (2003) investigated the puzzle by comparing anticipated and actual declines in spending at retirement. They use a non separable utility function in leisure and consumption. They found that a change in consumption comes as no surprise to most people when they retire. The average anticipated decline is larger than the actual decline. They also suggest that people might be more worried about the level of income which is anticipated at retirement age, instead of the level of consumption. Laitner and Silverman (2005) also use a non separable utility function in consumption and leisure and discrete work choices to specify a life-cycle model with these elements. They estimate different parameters in the model and most important the inter-temporal elasticity of substitution. The results show that whether consumption and leisure are substitutes or complementary determines whether people wish to increase or decrease consumption when leisure increase.
Schwerdt (2005) used German data and concluded that the drop in consumption is mainly found in the low income replacement group, where the drop is more than 30 percent, while the high income replacement group has an increase in consumption by more than 10 percent. The study therefore suggests that the fall in consumption differs between income replacement groups, and that the fall is not observed for all groups.
The aim of this paper is to analyze consumption- and income paths of the retired population, and to determine to what extent the life-cycle model fits the data for Norway.
Modeling was done for a particular group of Norwegian households that was tracked through their retirement transition period. For each household we observe pension income, labor income, wealth accumulation, saving and consumption. Each household contains one individual from the 1934 cohort. Data has it’s origin in different register files where each individual had a personal identification number code. The code makes it possible to link spouses and families together. The utilized files come from the tax offices and official registers containing demographic information. Data construction and programming was done in SAS system for Windows (Version 9).
The results show that the households analyzed increase their after tax income and consumption after one spouse has retired, and moreover they have a high level of net financial wealth. In addition it is found a connection between having a high income replacement ratio and having a low income before retirement, and those households who have high income replacement ratio increase consumption more than the households who have low income replacement ratio and a high income before retirement. The results also suggest that retirement decisions are independent of other decision making such as consumption and saving, since a high degree of early retirement and disabled are found in the sample analyzed at the same time as income and consumption is increasing for the sample. Independent of retirement behavior are people able to sustain their welfare. | nor |