FURTHER RESEARCH ON THE NATIONAL SWEDISH VALUE OF TIME STUDY

The analysis snow that the difference in value of time between willingness to pay and willingness to accept is due to resistance to pay more money and a difference in variance which ca be eliminated in modeling work. The analysis also show that it is possible to estimate 2nd order terms as well as 3rd order terms and that the inclusion of socio-economic variables is motivated and reduces the effect of 2nd order terms. In addition, the results indicate that in stated preference data the standard deviation of the value of time, and hence the confidence interval, is underestimated by about 50 percent if regard is not taken to the fact that the choices are not independent.