Commuting in multinodal urban systems: An empirical comparison of three alternative models



11. Conclusions

This article examined two questions. The first was to what extent the basic monocentric
urban model explains commuting distances within European urban areas. Is there a difference
between the commuting distance calculated on the basis of this model and the actual
commuting distance? The analysis shows that less than 4 % of the actual commuting distances
is explained by the concentrated urban model. So the answer to the first question is negative.

As the basic model starts from restricted assumptions regarding the direction of
commuting and commuter behaviour, a large part of the failure of the model, can probably
traced back to these assumptions. Therefore, the second question concerns the possibility of
adjusting the basic model in such a way, that it is more similar to the population's actual
behaviour. This would lead to a reduction in wasteful commuting and by this in a higher level
of explanation of the adapted urban model. Two alternatives were analysed for that purpose.
The first starts from the deconcentration of employment at locations on the radial of the
residential locations towards the centre of the city region. In this case only a slight
improvement in the explanation was achieved. Leaving aside the presuppositions regarding
the direction of commuting, and considering polynodality and cross commuter traffic instead,
lead to a second alternative. The results of this cross-traffic model show, that this is the case,
indeed, the share of wasteful commuting or, in other words, that part the other models could
not explain, decreased considerably. Adjustment of the model made the explanatory degree go
up from 4% to 40 to 55%. Thus the answer to the second question is affirmative.

Although the cross-model results in a considerable increase in the explanation of the
commuting distance, a large part still cannot be explained. This is partly due to
misspecifications of the density functions for employment and population. With respect to
the latter it would be desirable to use data on the actually working population. For, a part of
the potential labour force is, for reasons of for example further studies, unemployment or
incapacity, not actively working and by this not commuting. A further restriction to the
actively employed will also present a clearer picture regarding the decision to commute.

Another potential improvement of the explanatory level is the specification of
distance. In the models presented here distance was defined as the shortest road distance.
However, commuters possibly do not act in relation to this distance, but more likely to the
time spent on commuting. Unfortunately, we did not (yet) dispose of data on actual (or
perceived) travel time, but it is our impression that the inclusion of this in modelling would
indeed increase the explanatory power. Moreover, commuter behaviour is not just affected by
the costs of housing and commuting as included in the models discussed. Other aspects are
important too, like age, education, household stage, living environment, availability of
housing, the presence of certain kinds of transport and governmental policies. As these
aspects affect different groups of commuters in different ways, analyses of urban commuting
behaviour need to consider this heterogeneity of labour supply (see Cervero & Wu, 1997).

Related to this is also the misspecification of employment location. Two aspects are
crucial in this. Firstly, there is the assumption of exogeneity of employment to population
location. Several studies suggest, however, that present urban employment location
increasingly becomes endogeneous to population (Simpson, 1987; Giuliano & Small 1991;
Boarnet 1994). Not only do people follow jobs but also do jobs follow people. Secondly, in
addition it is suggested that, as with population, heterogeneity of employment is important
(Thurston & Yezer 1994). This implies that estimates of employment location decisions, and
whether the latter should be modelled exogeneous or endogeneous, differ by type of industry.
It is this heterogeneity in both employment and residential location and its spatial separation

22



More intriguing information

1. The name is absent
2. The Tangible Contribution of R&D Spending Foreign-Owned Plants to a Host Region: a Plant Level Study of the Irish Manufacturing Sector (1980-1996)
3. The name is absent
4. The name is absent
5. Towards Learning Affective Body Gesture
6. he Virtual Playground: an Educational Virtual Reality Environment for Evaluating Interactivity and Conceptual Learning
7. Disturbing the fiscal theory of the price level: Can it fit the eu-15?
8. The name is absent
9. Does adult education at upper secondary level influence annual wage earnings?
10. Modelling the health related benefits of environmental policies - a CGE analysis for the eu countries with gem-e3
11. The name is absent
12. Palvelujen vienti ja kansainvälistyminen
13. Urban Green Space Policies: Performance and Success Conditions in European Cities
14. Developmental Robots - A New Paradigm
15. Trade Openness and Volatility
16. The effect of globalisation on industrial districts in Italy: evidence from the footwear sector
17. An Economic Analysis of Fresh Fruit and Vegetable Consumption: Implications for Overweight and Obesity among Higher- and Lower-Income Consumers
18. Økonomisk teorihistorie - Overflødig information eller brugbar ballast?
19. The name is absent
20. Towards a framework for critical citizenship education