scope of econometrics { from a mathematical point of viewscope of econometrics { from a mathematical...
TRANSCRIPT
Econometrics Master Course: Methods 1. Econometrics in General
Scope of econometrics – from a mathematical point of view
Martin Treiber 1 of 10
Econometrics Master Course: Methods 1. Econometrics in General
General procedure of an econometric analysis
Martin Treiber 2 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Information flow of an econometric model
Martin Treiber 3 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Linking
Martin Treiber 4 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Chaining and feedback
Martin Treiber 5 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Models of time evolution (“dynamic models”)
I Special case of chaining: The endogenous variables at time t arethe exogenous variables at the next time step t+ ∆t
I The model itself is generally the same in all steps
I Sometimes, however, it has time dependent parameters(non-autonomous model)
Martin Treiber 6 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Models of time evolution (“dynamic models”)
I Special case of chaining: The endogenous variables at time t arethe exogenous variables at the next time step t+ ∆t
I The model itself is generally the same in all steps
I Sometimes, however, it has time dependent parameters(non-autonomous model)
Martin Treiber 6 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Models of time evolution (“dynamic models”)
I Special case of chaining: The endogenous variables at time t arethe exogenous variables at the next time step t+ ∆t
I The model itself is generally the same in all steps
I Sometimes, however, it has time dependent parameters(non-autonomous model)
Martin Treiber 6 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Models of time evolution (“dynamic models”)
I Special case of chaining: The endogenous variables at time t arethe exogenous variables at the next time step t+ ∆t
I The model itself is generally the same in all steps
I Sometimes, however, it has time dependent parameters(non-autonomous model)
Martin Treiber 6 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Application: Calculating the external costs of road traffic
Martin Treiber 7 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Model for limited growth
Limited growth according to the solution of the differential equation dydt =
1τ
(1 − y(t)
ys
)for the initial value y0 = 3 at the moment in time t0 = 1950
and the model parameters growth time constant τ = 10 and saturation
ys = 60. The result might represent the penetration rate for passenger
cars per person in %.Martin Treiber 8 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Structure of a modal split model
I Exogenous variables xkj : influencing factors j for mode k
I Endogenous variables yk: frequency of utilization for mode k
Martin Treiber 9 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Structure of a modal split model
I Exogenous variables xkj : influencing factors j for mode k
I Endogenous variables yk: frequency of utilization for mode k
Martin Treiber 9 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Structure of a modal split model
I Exogenous variables xkj : influencing factors j for mode k
I Endogenous variables yk: frequency of utilization for mode k
Martin Treiber 9 of 10
Econometrics Master Course: Methods 1. Econometrics in General
Mode choice for two alternatives:by bike or by public transport (PT)
Age Sextimeneededbike
costsbike
totaltraveltimePT
costsPT
choicebike
choicePT
Variables x1 x2 x3 x4 x5 x6 y1i y2iPerson 1 30 w 20 min 0e 30 min 1.00e 0 1Person 2 24 m 11 min 0e 20 min 2.00e 1 0Person 3 27 m 34 min 0e 15 min 2.00e 0 1...
......
......
......
......
......
......
......
......
......
......
......
......
......
These data can be obtained from interviews.
Martin Treiber 10 of 10