Low, Hamish Wallace; (1999) Simulation methods and economic analysis. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text out.pdf Download (12MB) | Preview

Abstract

There are two parts to this thesis: the first analyses the joint determination of saving and labour supply under uncertainty, using numerical dynamic programming to model life-cycle behaviour. The second uses simulation to analyse the problem of equilibrium selection when individuals are myopic and interaction between individuals is predominantly local. The value of using simulation methods is that this enables the analysis of more complex and realistic models than is analytically feasible. In the first part, individuals choose consumption and labour supply in each period of their lives, without knowing what their future wage rate will be. Existing literature treats these choices separately: for example, modelling consumption choices holding labour supply fixed. The uncertainty over wages leads individuals to reduce their consumption and to work longer hours when young in order to accumulate assets to use in case of a low wage in the future. However, consumption is smoother than in models with an exogenous labour supply. Further, since labour supply is flexible, individuals can increase their leisure when the realised wage rate is low, and this reduces the welfare cost of uncertainty. The model is extended in two ways: first, household choice rather than individual choice is modelled, thus giving two sources of income and providing a degree of insurance to the household (the extent of this depends on the correlation between the wages of the primary and secondary earners); and second, social security is introduced and this provides explicit insurance, but also discourages participation. The simulations show how labour supply and participation choices change across the life-cycle and how these choices interact with the savings decision. In the first part of this thesis, individuals are assumed to be rational, in the sense of using information optimally, but the second part models behaviour in a more general setting if the high cost of computing the optimal strategy leads individuals to follow rules of thumb. There is a wide literature showing that rule-of-thumb behaviour can lead to the selection of particular equilibria, but in these models the expected waiting times to reach these equilibria are sometimes fantastically large, particularly as population size increases. This part of the thesis shows the reduction in this expected waiting time caused by making interaction local and by making strategy payoffs stochastic. The first innovation here is to combine these two effects, and the second innovation is to show the extent of skewness and variance in waiting times.

Download activity - last month

Export as XMLCSVJSON

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item