Modelling the evolution of heterogeneity in residential mobility

Area of Science:

• Econometrics
• Sociology
• Demography

Background:

• Migration probabilities vary within populations (heterogeneity) and change over time (nonstationarity).
• Existing models often struggle to capture both individual differences and temporal dynamics simultaneously.

Purpose of the Study:

• To develop a flexible model for migration probabilities accounting for heterogeneity and nonstationarity.
• To enable comparison of heterogeneity at different time points.
• To model individual tastes influencing migration decisions.

Main Methods:

• Representing heterogeneity using a self-binned (SB) mixing distribution.
• Modeling nonstationarity with arbitrary logistic scaling functions.
• Calibrating the model with existing data on older renters.

Main Results:

• The proposed model effectively captures both population heterogeneity and temporal changes in migration probabilities.
• The model fits well with previously published data.
• Graphical representations illustrate the evolution of heterogeneity over a decade.

Conclusions:

• The SB mixing distribution and logistic scaling functions provide a robust framework for analyzing dynamic migration patterns.
• This approach allows for detailed insights into how individual migration behaviors evolve and differ over time.