| Authors: | Lasse Fridstrøm |
| Report nr: | 1142/1999 |
| Language: | Norwegian |
| Attachments | Sammendrag |
The municipalities of Sweden will implement the "Vision Zero" of road safety at varying pace, by - e g - rebuilding the roads, lowering the speed limits, or enforcing safety standards in public agencies. This "natural experiment" might pave the ground for a powerful, pooled cross-section/time-series analysis of road accidents and their determinants, by means of which the practical impact of the "Vision Zero" should be estimable, with steadily increasing precision over time. The present "pseudo Monte Carlo" experiment examines under what conditions statistically significant effect parameter estimates can be expected from a generalized Poisson regression modeling approach as applied to municipal accident counts. Results are generally encouraging with regard to feasibility.
