Time and cost are the main variables in travel choice models. The microeconomic foundations of this type of specification lies ultimately on the so-called time allocation theories, which contribute to provide interesting interpretations of the so-called valuation of travel time savings (VTTS). This VTTS is the willingness to pay to diminish travel time by one unit. The most popular of these interpretations is the comparison with the wage rate, which follows the framework created by Becker (1965), later adapted to discrete travel choices by Train and McFadden (1978) in their goods-leisure model. A slight modification of the conditions assumed in this framework – namely a fixed working schedule - generated an expenditure rate specification (Jara-Díaz and Farah, 1987) which contributed to provide clear explanations for the lower values for the VTTS obtained for retired people when compared with active individuals having similar income, or the larger VTTS obtained for longer distances urban trips.

The interpretation of the VTTS is much richer when the underlying microeconomic framework rests on the theory developed by DeSerpa (1971), who departed from Becker’s approach by introducing work time in the utility function and a series of technical constraints linking time assignment and consumption, containing Becker’s result as a particular case. Under this view, the VTTS happens to be equal to the value of doing something else minus the intrinsic value of the marginal utility of travel time. The value of these other activities can be seen either through the total value of work – given by the wage rate plus the intrinsic value of time assigned to work - or the value of leisure. Until recently these values could not be calculated, but Jara-Diaz et al (2008) have presented a DeSerpa like framework - expanded from Jara-Diaz and Guevara (2003) - from which a model system of time use equations was derived and estimated using three different experimental data sets, obtaining analytical estimations of the values of leisure and work.

In this paper this theoretical model system is explained and applied to a large time use data base constructed from the last origin –destination survey available in Santiago, Chile, which is representative of the population in that city. It was conducted from July 2001 until April 2002 (DICTUC, 2003). From the travel information contained in the household surveys we were able to construct 9464 weekly observations where six aggregated zones of the city were adequately represented. The procedure to generate weekly time use observations from daily travel diaries is described in detail in Olguin (2008); this was supplemented with expenditure data from other sources. Data reveals marked differences by geographical sector (associated with income), by gender and age. For example, people from the (richest) East sector dedicate less time to work in average. On Sunday, people from the Centre are the ones who work more. On week days and Saturday those from the other zones (North, West, South and South East) work more; they also dedicate more time to travel in all types of days. In general, older people spend more time at home, less time at work and more time to leisure activities during the weekend.

The large number of individual weekly work-leisure cycles was enough to create at least six aggregated segments: youngsters, elderly and four sub-groups for individuals within the range 25-64 years old, sub-dividing by gender and income (zone). We estimated the time assignment model system for these groups using maximum likelihood techniques. We report the values of work and leisure, which happen to be negative for the two male groups, larger in absolute value for the richest one but quite similar regarding their values relative to the corresponding wage rates. Both female groups exhibit positive values of work, somewhat larger for the richest sub-group but smaller when corrected by income. The values of leisure relative to income are quite similar within each gender, but larger for women than for man. When examining the effect of age differences, the results of the time use model system suggest that the displeasure for work increases with age. Explanations are offered for these empirical findings based upon the socio-economic characteristics of each segment. This is the first large scale application of the new theory to calculate the values of time that lie behind the VTTS.

Becker, G. (1965) A theory of the allocation of time, The Economic Journal, 75 493-517.

DeSerpa, A. (1971) A theory of the economics of time, The Economic Journal, 81 828-846.

DICTUC (2003). Actualización de encuestas Origen Destino de viajes, V Etapa, Informe Final a Sectra, Santiago.

Jara-Diaz, S.R. and Guevara, A. (2003) Behind the subjective value of travel time savings: the perception of work, leisure and travel from a joint mode choice - activity model, Journal of Transport Economics and Policy, 37 29-46.

Jara-Diaz, S. R., M. Munizaga, P. Greeven, R. Guerra and K. Axhausen (2008) Estimating the value of leisure from a time allocation model. Transportation Research B 42, 946-957.

Olguin, J. (2008) Modelos de uso de tiempo para el Gran Santiago (Time use models for metropolitan Santiago). M.Sc. Thesis in Transportation, Departamento de Ingeniería Civil, Universidad de Chile.

Train, K. and McFadden, D. (1978) The goods / leisure trade-off and disaggregate work trip mode choice models, Transportation Research, 12 349 - 353.