I have been looking at the evolution of assets in LifeCycleModel41 (experienceasset (Female Labor Force Participation History)). What is not intuitive to me is that the Mean accumulated asset value remains near zero (after being spent down at age 32) until age 40. Surely if household can accumulate some assets by age 27, then post childbirth workers aged 32-40 should be saving *something* visible on the graph.
Are children really so distracting that their marginal attention needs overwhelm asset accumulation until the very last possible moment of the optimal asset accumulation strategy?
1 Like
The issue is that the parameters in all the Intro to Life-Cycle Models are rather arbitrary (deliberately). Hence they are not chosen to give sensible model results. The focus is instead just on setting up models and solving them.
If you look at the paper Attanasio, Low & Sancez-Marcos (2008) you will see a similar model, but with sensible numbers. You can find code implementing this model in VFI Toolkit here:
In short, the weird behaviour is just because of not using ‘realistic’ parameter values. It is the correct behaviour for the model with those parameter values.
I have the vague idea that at some point I will create an ‘Intermediate Life-Cycle Models’ which will go through the steps of calibrating/estimating models in a sensible fashion, and doing some counterfactual exercises. But that remains a dream for now.
1 Like