The travel cost of getting from your home to work includes much more than just the price of petrol or the bus fare, write Dr Kevin Credit, Assistant Professor at MU's National Centre for Geocomputation and Dr Conor O'Driscoll of University of Groningen.

The relationship between how you travel and the environment you live or work in is well documented in academic literature. Usually, this relationship explores how different built environment characteristics are associated with different patterns of travel behaviour, thereby providing an evidence base for policymakers to explore the merits of different land-use and transport policies.

But our research takes a different approach. We analyse how these environments impact travel costs for commuters living and working in the Dublin Metropolitan Area. In this case, the "costs" that we calculate include a monetised estimate of everything that the user must pay when using a mode.

For example, this includes both the cost of petrol or the bus fare and also the opportunity cost of your travel time (adjusted to account for differences in comfort of various transport modes) and per-km ownership and maintenance costs.

Our data consist of "flows", that is, a dataset containing the trip origins (residential electoral divisions) and trip destinations (workplace electoral divisions) of commuters. We then transplant these flows to relevant transport networks (i.e., roads for car users, bus routes for transit users etc), which allows us to account for factors like congestion levels, infrastructure characteristics, and, for public transport, service frequency along different flows/routes.

Most of the data used in this analysis comes from the 2016 Census while the remainder comes from Google MapsTransport For Ireland, the Revenue CommissionersOpenStreetMap, and the Residential Property Price Register. Using this data, we employ random forest modelling techniques to measure how mobility costs differ across modes depending on built environment and transport network characteristics in residential and workplace electoral divisions. By including individual-specific variables, we can start to isolate the effect that built environments have on the costs associated with travelling by specific modes, along specific routes. More technical details about this analysis can be found here.

Of all the variables included in our models (one model per transport mode), six to 10 of the most important variables influencing costs relate to built environments and congestion-effects. This vindicates a long-disputed maxim arguing that, although individual measures of built environment characteristics (i.e., population density, land-use mixing) may have a limited effect on promoting sustainable travel, their overall effect can be substantial.

While this alone has clear implications for land-use policy, our results go further and our models allow us to show non-linear relationships in this data. For example, active travel becomes cheaper as intersection densities increase, but it starts to become more expensive beyond a certain point. Increased densities reduce the distance people need to travel when commuting but, after a certain point, areas may become "too dense" or networks may become "too busy", for existing infrastructure to handle, making active travel relatively unattractive. Interestingly, the initially positive effects are not fully cancelled out by these latter negatives, suggesting that, overall, density-increasing measures reduce the cost of travelling actively.

We also find that the costs of using public transport are sensitive to congestion effects, particularly in dense areas, which is a bit surprising. Generally, one of the benefits of public transport is that each additional person using the mode (up to a point) doesn't make the trip any slower; whether there are 20 or 50 people in a train car, the train will move at the same speed. This is not the case with cars, where each additional driver slows trips along that route due to congestion.

Interestingly, it is also not the case with public transport in Dublin. This is likely due to the fact that Luas, buses, and cars all share the same infrastructure in the capital. Shared infrastructure implies shared congestion, meaning that other factors, like comfort and convenience, may become especially important in determining mobility costs; and nothing is as comfortable or as convenient as cars in the current Irish landscape. If these modes face the same congestion, the cost of using public transport for short distances may be higher than those for cars, creating a "crowding out" effect. This may partially explain patterns of rising car-use for journeys shorter than 15-minutes across Ireland.

However, it may also partially explain our finding that mobility costs for public transport are relatively high in some parts of the inner-city – the dense parts of the city where public transport should be most effective (if it were grade-separated). We find that public transport becomes relatively expensive in traditionally working-class neighbourhoods like Inchicore, Dolphin’s Barn and Kimmage, and also Phibsborough on the north side. This leads us to conclude that congestion and network effects may influence mobility costs for public transport more than built environment effects. Given where these cost increases occur, we also suspect that potential social/political inequalities may play a role.

Our other results paint familiar pictures. Walking becomes more expensive in, and between, suburbs like Kimmage, Clonskeagh, Cabra, Finglas, Donnycarney and Ballyfermot. Cycling becomes more expensive in and around Dublin Airport (including Turnapin, Balgriffin, and Kinsaley), Donnycarney, Coolock, northern Lucan, Palmerstown, Tallaght-Belgard, Monkstown, Dalkey and Shankill.

Car use is most expensive in inner-city areas, areas along the M50 motorway (e.g., from Palmerstown to Castleknock) and in places like Ashtown, Cabra, Donnycarney and Blackrock. Aside from some inner-city areas, public transport becomes more expensive in suburban areas like Palmerstown, Castleknock, Donnycarney and Cabra.

If a move towards more sustainable travel is an objective, measures must be in place to reduce the costs associated with this transition. In this context, costs influence "transport hierarchies", something we understand to reflect the number of "viable" travel options available to individuals. If changing mode-specific costs creates shifts in this hierarchy, it provides a policymaking mechanism for stimulating individuals to transition towards more sustainable travel.

But what does this mean for the Dublin Metropolitan Area? For this research, we compared the average costs of travelling 15km at the morning peak in Dublin with those in Chicago. This exercise showed similar trends in the overall results (i.e., public transport is the cheapest mode in the hierarchy and walking is the most expensive, due mostly to the fact that walking is very slow, and thus presents greater opportunity costs). But it also highlighted how cycling and public transport use is actually cheaper in Chicago than in Dublin, possibly indicating that the efficiency of Dublin’s public transport and cycling networks are lagging behind those of global cities.

Average total cost of travelling 15km at the morning peak in Dublin and Chicago 

Mode Dublin Chicago
Auto €13.26 $9.94
Public transport €10.70 $8.71
Cycling €16.84 $9.39
Walking €54.26 $67.89

So, how can we reduce the costs associated with travelling more sustainably? Specific measures, like segregated bus and cycle lanes are mechanisms which can reduce the costs of using public (and active) transport by reducing congestion-effects and increasing network connectivity – costs which could be reduced further if these networks are extended. The construction of an underground Metro system would also decouple public transport from street-level congestion, making public transport more attractive and useful overall.

But these measures must be implemented in a spatially coherent manner, such that shifts in local transport hierarchies can be realised. A good example of this is captured by the various Bus Connects schemes being rolled out across Irish cities. However, we caution that these efforts may be hindered unless land-use policies are configured to play a complementary role. Decades of uncoordinated urban sprawl across Ireland have created the conditions for existing, car-dependent transport hierarchies to emerge, meaning greater collective efforts may be required to reverse these trends.

This piece originally appeared on RTÉ Brainstorm