Whither the use of economics in transport policy? In mobility, as in other areas of life, people do not always act in their own interest, and this reduces the effectiveness of transport policy instruments such as congestion pricing. In this contribution, we discuss some approaches that can be used to improve the effectiveness of transport policy, such as using the power of social networks to encourage more sustainable behaviour, introducing temporary changes in order to help people breaking habits, reframing travel information to increase the awareness of the full cost of travel choices…. 01. Behavioural economics and policy support: generalities Current approaches to transport policy evaluation are often firmly embedded in a neoclassical economic framework, which assumes perfect rationality in all decision making. However, this intellectual framework has been questioned by an alternative approach, behavioural economics, which aims to align economic analysis with insights from other behavioural sciences, such as psychology. Behavioural economics is relevant for transport policy makers for at least three reasons: (a) it undermines several dimensions of the approach that is used to assess the costs and benefits of transport infrastructure (b) it challenges the techniques that are used to predict travel behaviour (c) it raises deep questions regarding the effectiveness of policy instruments. These three problems are all important, but in the current text we focus on the implications for the design of transport policy instruments. Before proceeding, we should note that, for policy makers and their advisors, there are essentially three possible approaches to deal with the findings of behavioural economics (Gifford and Checherita-Westphal 2008): Ignore the problem. The first approach is to ignore the problem or to treat it marginally. Proponents of this approach point out for instance that findings in behavioural economics are often based on experiments that haven taken place in laboratory settings. It is possible that the deviations from rationality follow from the artificiality of the laboratory setting, and do not reflect how people act in real life. In the lab, the financial stakes of the decisions taken by the participants tend to be low, and the choice problems are often artificial. For instance, in transport problems, the choice set is often limited to 2 routes or 2 modes (car and bus, for instance). It is then possible that apparently irrational behaviour follows from the fact that people are not familiar with this type of decision making, and do not have the time to learn from their mistakes. In real life, people have concrete and substantial stakes in correct decision making, and they feel the consequences of poor choices. A slightly less radical approach consists in acknowledging that people’s behaviour is not perfectly rational, but that the rationality assumption is good enough as an approximation of people’s actual behaviour to be useful in transport modelling. Some also argue that the errors implied by the rationality assumptions are minor compared to those following from relying on incomplete or incorrect data in transport modelling. In both cases, non-rational behaviour is ignored in modelling transport choices. Heuristics as the new norm for decision-making. The second approach is to accept that people are less than perfectly rational and to incorporate deviations from perfect rationality into modelling. In this approach, the norm is the use of heuristics, which is the technical term to refer to ‘any “rule of thumb” or simple rule of behaviour by which a person solves a problem’ (Cartwright, 2011, p. 27). Proponents of this approach refute that classical rationality should be the normative standard. They argue that, even if the results of heuristics are less than optimal, they result from the necessities of real life conditions: people do not have the time nor the information that is necessary to take fully optimal decisions, and it can be better to take a decision that is ‘good enough’ (for instance, always taking the same bus line at the same hour to go to work), rather than continuously reassessing the situation (for instance, by recalculating the shortest and cheapest route to work every morning). Libertarian paternalism. The third approach is to remedy the problem of imperfect rationality. This approach retains rational behaviour as the normative standard, but acknowledges the existence of non-rational behaviour. The main policy implication is that policy-makers should implement measures that align individuals’ choice processes with their own best interest. This is the approach known as “libertarian paternalism”. Libertarian paternalism has largely been brought to the attention of policy makers through the work of Thaler and Sunstein (2008), which has drawn the attention of policy makers all over the world. It has also rapidly influenced actual policy making. In the UK, it has led to the creation of a Behavioural Insights Team, whose experiences have been described vividly by Halpern (2015). I will now further elaborate on this third approach. The central message of Thaler and Sunstein (2008) is that there are a lot of situations where, according to behavioural economics, people run the risk of acting against their own interest: they are over-optimistic about their own judgement, take both unnecessary risks and exaggerated precautions (depending on the situation), let their decisions be influenced by irrelevant information etc. Thaler and Sunstein refute the argument that the ‘anomalies’ detected by behavioural economists only hold in the lab, and give numerous examples taken from real life. The central policy innovation they propose is the concept of “libertarian paternalism”, which they define as policies that: maintain or increase freedom of choice – the “libertarian” side. try to influence choices in ways that will make choosers better off, as judged by themselves – the “paternalistic” side The central concepts in their approach are “choice architecture” (the organisation of the context in which people make decisions) and “nudges” (small features designed in the environment of choice making). Libertarian paternalism and transport policy Most applications of the “nudging” approach proposed by Thaler and Sunstein (2008) are examples of situations where people do not act in their own interest. In contrast, most transport problems (congestion, air pollution, accidents) arise because private and public interests are not aligned – in economists’ jargon, because there are externalities. One may thus wonder what the relevance of this approach is for transport problems. The answer is threefold. First, in some cases, externalities are exacerbated because people make choices against their own interest. For instance, there may be instances where people would be better off taking public transport rather than a car, but still take their car. This is usually attributed to ‘status quo bias’: people tend to stick with the current situation, even if they would gain be changing -we will discuss possible causes further in this paper. Second, behavioural biases such as the status quo bias may hinder the effectiveness of some policy instruments such as congestion charging: even if the congestion charge is set high enough to make a switch to public transport the optimal choice, people may still stick to using their cars. This could imply that a congestion charge that would induce an important behavioural change should be set a level that would not be politically feasible. An alternative approach would be to overcome status quo bias using other policy instruments that would complement congestion charging. Third, the insights derived from the “nudging” approach can help to improve the effectiveness of market based instruments and regulation, but also of so-called “soft” policy measures, such as (sustainable) travel plans, promotion of car sharing, or leveraging social media or other forms of peer pressure (Avineri, 2012). We will illustrate this with some concrete examples. 02. Social influences and travel behaviour Aristotle already wrote that man is, by nature, a social animal. Some minimal introspection is sufficient to understand that social aspects are key in several mobility related decisions, including the decision to travel in the first place. For non-economists, it may therefore come somewhat as a surprise to learn that, in the traditional economic models of travel behaviour, people are considered in isolation. Their decision to travel, the timing of their travel, the transport modes they use… all these crucial decisions are assumed to depend on the individual’s ‘preferences’, but there is no (at least no explicit) discussion of how these tastes have been shaped by, for instance, the mobility needs of their families, the desire to meet people in their social networks or the travel habits of their peers (Avineri 2012). Fortunately, this is changing now, and economists are increasingly working on how social interactions shape travel behaviour. I will here summarize the key insights. A first reason why people’s mobility behaviour is affected by others is simply that decisions are sometimes made at the level of a group (Chorus 2012). For instance, in the context of a household, parents do not just travel from home to work, but also bring children to school or day-care (and fetch them there as well), do some shopping for the household, drive children to hobby activities etc. The recognition of these household interactions is the subject of the so-called Activity Based Modelling approach to travel modelling (see Rasouli and Timmermans 2014). Second, people tend to compare with others when taking decisions or evaluating the consequences of these decisions. Abou-Zeid and Ben-Akiva (2011) distinguish the following three channels through which social comparison could affect behaviour: People obtain information from others, and this affects both their level of awareness of options and the perception of the attributes of these options. The satisfaction and advice of others is used in the assessment of the attractiveness of different options. This will especially be the case in fields where people have little experience: gaining information from other allows to spend less time and energy in finding the relevant information. In the case of transport, this is obviously an important element whenever an individual has to take decisions outside his daily routines (for instance, after changing jobs, after moving or when planning for holiday trips). People seek the approval of others and adhere to group norms. In other words, they exhibit herd behaviour, are influenced by peers, conform, etc. This can be both an important barrier to change, but also an engine for change if the group norm starts changing. Downward comparison (i.e. comparing oneself to others who are perceived as doing worse on an item of comparison) may make one feel happier, while upward comparison to others who are better off may make one feel less happy (Wills, 1981). All his raises several questions: How strong are social influences in shaping behaviour? Are these relevant from a policy point of view? How can we measure them? How do they affect our assessments of the welfare effects of transport policies? Social influences and conformism The answer to the first question is simple: social influences are very strong indeed in influencing behaviour. Psychological experiments show that (see Thaler and Sunstein (2008, Chapter 3)) even for very simple tasks, people tend to conform to the judgement of other people’s, including of complete strangers. The policy relevance of these issues is evident. For instance, people may stick with travel habits (such as commuting by car) even when there are faster and/or cheaper options available, simply because most people they know also commute by car. Social interaction and social learning/imitation may have a considerable role in responses to ‘soft’ transport policy measures (e.g. travel awareness campaign, travel plan, car-sharing). ‘Soft’ measures, may therefore be more useful if they have a local and personalised focus (e.g. targeting schools, companies, communities or other groups that bring a feeling of belonging), instead of aiming at the entire population (Sunitiyoso et al. 2011). The influence exerted by one’s social peers can also play a key role in the development of the critical mass that is needed for the financial viability of public transit, and can thus create virtuous cycles (Dugundji and Gulyas 2008). Of course, such a virtuous cycle can only materialise if the supply of public transport can accommodate the increase in demand. Social influences in transport are not limited to mode choices. For instance, Axsen and Kurani (2011) have studied car buyer perceptions of plug‐in hybrid electric vehicles (PHEVs), and concluded that conformity and other mechanisms of social influence played a role in the assessment and adoption of alternative fuel vehicles. Social norms and information obtained from colleagues are also key in the decision to telecommute (Páez and Scott 2007, Wilton et al. 2011, Scott et al. 2012). Other forms of social influence and conformity are less benign. For instance, Fukuda and Shigeru (2007) have found that, in Tokyo, commuters are more likely to park their bicycles (illegally) on-street when this also done by other people at the railway stations they visit often. Social learning and travel behaviour Learning does not only pertain to the behaviour of others, but also to factual information, or to a combination of both. For instance, Bartle et al. (2011) have observed how a small group of commuter cyclists interacted with one another through a map-based website, where they could share their routes and other cycling-related information. Besides its functional role in diffusing practical travel information, the information sharing process performed a social role as well: perceived in-group membership reinforced positive views of cycling as a commuter mode. Social interaction and the emergence of norms Sometimes, the roles are reversed, and it is the social interactions related to transportation activities that lead to the emergence of social norms. Consider slugging, an informal system of carpooling in the Washington, DC area. Slugging emerged when Virginia introduced High Occupancy Vehicle (HOV) lanes in the early 1970s as a tool to encourage carpooling. As a result, a practice developed where single drivers picked up riders alongside the road (slugs) in order to meet the requirements for driving in the less congested HOV lanes. Mote et Whitestone (2011) have explored the social practice of slugging, where they have observed a spontaneous emergence of informal institutions with its own norms, etiquettes and practices. From a policy point of view, the case of slugging represents a typical case of the “law of unintended consequences”. Whilst the objective of the HOV policy was to encourage car-pooling, the result was that it took ridership away from public transit. Social interaction as a travel motive Social networks do not just affect travel choices once the decision to travel has been taken: they are a key cause of travel, especially in the case of leisure travel. To understand someone’s travel behaviour, it is thus important to know the composition and the geography of his social network (Axhausen 2008). Social influence and welfare assessments The existence of social influences is also relevant for the evaluation of transport project and policies. Indeed, the metrics that are used in economic appraisals typically assume that an individual’s welfare does not depend on the goods and services owned by others, or by the welfare of others. However, Abou-Zeid and Ben-Akiva (2011) have shown that commute (dis)satisfaction is influenced by social comparisons: people may feel less bad about long commuting times if they realise that several people in their peer group face even longer commutes. The benefits of infrastructure projects that aim at reducing commuter times may thus be less than expected if satisfaction levels adapt to the new context. 03. Status quo bias and travel behaviour People tend to stick with the status quo. They like it so much that they often forego opportunities for improving their situation, without even giving consideration to these opportunities. This behavioural oddity has been found to be true in a wide range of situations, ranging from the choice of their energy provider to organ donations. A special case of ‘status quo bias’ arises when the supplier of a good or a service provides a default option (for instance, for insurance schemes or mobile phone subscriptions): it has been found that people mostly stick with the proposed default. Behavioural scientists have come up with several possible explanations for this phenomenon. For instance, it may simply be that when defaults are proposed, people feel that the default options are “endorsed” (Thaler and Sunstein 2008, p 35), and thus probably superior to the alternatives. Choosing the default saves them the effort of investigating the other options in detail. In several situations (for instance, when choosing the settings of a newly installed software), this may indeed be the wisest approach. In more general terms; it may be that the cost of searching for new alternatives, and understanding them, is too high compared to the possible benefits of switching. Reusing past solutions makes life easier and less risky (Garling and Axhausen 2003). Whenever people develop habits, this allows them to save on cognitive efforts. Not surprisingly, inertia has also been observed in mobility behaviour. Once people have chosen a transport mode and a route to reach a certain destination, they will stick to this mode (Gardner 2009) and route (Srinivasan and Mahmassani 2000). There are several reasons why status quo bias may be even more pronounced in mobility behaviour than in other human activities. Indeed, in the case of mode choice, people are also influenced by the symbolic and affective value of cars: people seem to choose on the basis of an ‘‘affect pool” associating a positive tag to car use. This “affect heuristic” could be the result of repeated mental associations over time, which result in people generating intuitive responses that could previously have been the outcome of analytic thinking (see Steg 2003). People also tend to underestimate the costs of cars because there is a time gap between car use and the payment of the costs. In other words, the variable costs of the car are not salient enough (Metcalfe and Dolan 2012). Innocenti et al.point to an important conundrum. The ‘status quo bias’ implies that economic incentives may be relatively ineffective in reducing people’s preference for cars. One may therefore be inclined to conclude that transportation policies should focus more on initiatives which increase individual awareness in making choices. However, until now, ‘‘soft’’ policies based on the provision of information also tend to achieve only modest results. Just like mode choice, route choice appears to be very stable through time, even when circumstances change. Several international studies based on field observations have shown that a substantial share of people do not usually take the shortest route. This has recently been confirmed by Di et al. (2014). After the sudden collapse of I-35W Mississippi River bridge in Minneapolis on August 1, 2007, people had no choice but to start using other routes. However, Di et al. found that, once the replacement bridge was in place, it saw less traffic than the original bridge, even though it provided substantial travel time saving for many travellers. Di et al. therefore propose the hypothesis that, once people have chosen a route, they will only switch routes if the travel time on the chosen routes are outside some threshold from the shortest ones. As far as policy is concerned, Garling and Axhausen (2003) argue that it is important to understand how habits are broken, in other words, how to ensure that choices become deliberate and rational again. Several approaches to ‘habit breaking’ have been proposed, some more practical than others. The most promising idea in these times of ever more reliable and immediate travel data, is to provide costless multimodal travel information. If the information is considered reliable, this may slow down the emergence of inertia (Chorus and Dellaert 2012). The potential of multimodal travel information to prevent people from taking suboptimal decisions has improved significantly with the use of crowdsourced travel apps. Those apps combine real-time information on the state of the transport network with historical data on the habitual travel choices of a user, to warn the user that he should deviate from his routine, for instance because of a disruption in the transport network. Another possibility is for policy makers to introduce a temporary structural change, such as offering auto drivers a temporary free bus ticket. Fujii and Kitamura (2003) have conducted an experiment in which a one-month free bus ticket was given to 23 drivers in an experimental group but not to 20 drivers in a control group. The results showed indeed that attitudes toward bus use became more positive and that the frequency of bus use increased after the intervention. 04. The framing effect and travel behaviour When economists discuss how people make decisions on the transport modes they will take, or the routes they will follow to reach a given destination, they usually assume that people only consider the contents of the information which is provided. For instance, it should not matter whether your route planner tells you that route A is 10 kilometres longer than route B, or that route B is 10 kilometres shorter than route A. These are just two different ways to say the same thing, and it seems like simple common sense that people will act the same as well in both settings. Except that they don’t. Research in numerous field has shown that people’s decisions are affected by elements that are, objectively speaking, irrelevant. For instance, in a medical context, Wilson et al. (1987) have conducted psychological experiments where people were confronted with the decision whether a hypothetical patient described as their ‘father’ should return from the intensive care unit to the regular floor. It turns out that people are more willing to endorse this return if his chances of surviving were presented as 90 per cent than when told that his chances of dying were 10 per cent. Thus, although’90 per cent chance of surviving’ is objectively the same as ’10 per cent of dying’, people were sensitive to the ‘framing’ of information. Choices thus depend, in part, on the way in which problems are stated (Thaler & Sunstein, 2008). And this has direct implications for transport policy. Some travel planners currently do not just provide information on travel distance and estimated travel time, but also on the environmental implications of travel choice. Suppose, for the sake of concreteness, that travel planners would not just report the different travel options, but would also compare the CO2 emitted in one option with emissions for two other options. There are at least two ways to frame CO2 emissions: as ‘losses’ compared to the option with the lowest emissions or as ‘gains’ compared to the option with the highest emissions. Avineri and Waygood (2013) have conducted an experiment where people were confronted with the CO2 emissions of three travel options for a five-mile trip: a bicycle, a car occupied with four passengers, or a single occupancy four-wheel-drive vehicle. The participants were asked whether the differences in CO2 emissions were perceived as ‘‘about the same,’’ ‘‘slightly different,’’ or ‘‘much different’’. The results obtained by Avineri and Waygood confirmed that people are more likely to label the alternative options as “much different” if the information emphasizes the possibility to reduce environmental damage compared to the worst travel option. Of course, one may object that people’s evaluation is also affected by the symbolic value of travel options such as cycling and four wheel drives. For “deep greens”, for instance, it is likely that four wheel drives generate negative feelings, independently of the actual amounts of CO2 that are emitted. To avoid this effect, Avineri and Waygood did not describe the actual transport modes: they just reported the differences in CO2 emissions. Thus, the differences in perceptions cannot be due to positive or negative feelings associated with the transport options. The framing of the environmental implications of travel choices can thus to influence these choices, at least if people do actually care about the environmental impacts of their travel choices. Avineri and Waygood emphasize that information provision in itself is unlikely to result in substantial behavioural changes. However, environmental information can be embedded within broader packages of incentives for sustainable travel behaviour, including ‘soft’ measures. It is then important that such ‘soft’ measures are based on a rigorous analysis of how the framing of the measures are likely to affect behaviour. Another example is the ‘miles per gallon’ illusion: in the US, fuel efficiency is expressed as miles per gallon (rather than as litters per 100 km). Larrick and Soll (2008) find that people systematically misunderstand the concept of miles per gallon (MPG), and tend to undervalue small improvements on highly inefficient vehicles. The reason is that, instinctively, people tend to see the amounts of fuel consumed by a vehicle as a linear function of its MPG. In reality, this function is hyperbolic, and this has far-reaching implications. Let us follow the example given by Larrick and Soll, and consider a car who drives 10,000 miles. A car with 12 MPG consumes 833 gallons over this distance, while a car with 14 MPG consumes 714 gallons. This corresponds to a total fuel savings of approximately 120 gallons. Now compare this with the gain obtained when a car with 28 MPG is replaced by a car with 30 MPG: the total savings is barely 23 gallons (357 gallons – 333 gallons). Clearly, the greatest environmental benefit is to be obtained by focusing on replacing the gas guzzlers, although the change in MPG is the same for both cases. If the standard is expressed as “gallons per mile” instead, consumers understand better how much fuel they are using on a given car trip or in a given year. Even if these findings are only directly relevant for the US, they do point to a broader lesson: if we want people to make better informed decisions, it is essential that this information is expressed in terms that they can understand without additional intellectual effort, and that they can translate immediately in terms that are directly relevant for them. In this case, the total amount of fuel one consumes over a year is more directly relevant than the distance one can cover with a given amount of fuel. Some other examples Next to the cases we have explored above, there are other tools and insights from behavioural economics that are useful for transport policy, for instance: Increasing the salience of the cost of the variable costs of a car (for instance, by providing information on the life cycle costs of the car) could compensate some of the behavioural biases that induce people to favour cars over alternative transport models. Actually, Thaler and Sunstein have concluded that the “most important modification that must be made to a standard analysis of incentives is salience.” Route planners (especially those supplied by local authorities for recreational travel) could propose “sustainable” travel modes as the default option (while still leaving open the option of providing route advice for car trips) (Avineri, 2012) – it is one of the key insights in behavioural economics that defaults have a strong impact on the options people choose. If people wish to conform to the opinions and values of their peers, then the use of social networking (for instance in combination with car sharing, workplace and school travel plans) could encourage modal shift (Avineri, 2012). Moreover, people turn to their social networks to learn from others’ experiences when making travel decision, including strategic decisions such as telecommuting. An interesting tool to help people with decisions that involve difficult trade-offs, is the use of ‘collaborative filtering’: this tool (which has become famous because of its extremely effective use by Amazon and Netflix) provide recommendations to people, based on the past choices of people who are ‘similar’ to them. Thus, by no means does behavioural economics entail that market based instruments and other forms of traditional policy intervention become obsolete. Rather, the insights from behavioural economics can help to improve the impact of such measures. Laurent Franckx Laurent Franckx holds a PhD in economics from the University of Leuven. He is currently affiliated with the Belgian Federal Planning Bureau as an expert in Energy and Transport. Over the last two decades, Laurent has worked in diverse positions ranging from academia to consultancy firms, and has worked on a broad range of topics in the fields of environmental and transport economics. His recent work has focused on applications of behavioural economics to transport policy, and the implications of new transport technologies and business models (such as self driving vehicles and shared mobility) on long-term developments of the transport sector. Abou-Zeid, M. & M. Ben-Akiva. The effect of social comparisons on commute well-being Transportation Research Part A 45 (2011) 345–361. Avineri, E. (2012), On the use and potential of behavioural economics from the perspective of transport and climate change. Journal of Transport Geography 24, 512-521. Avineri, E. and E.O.D. Waygood (2013), Applying valence framing to enhance the effect of information on transport-related carbon dioxide emissions, Transportation Research A 48, 31–38. Axhausen, K.W. (2008). “Social networks, mobility biographies, and travel: Survey challenges.” Environment & Planning Part B, 35, 981-996. Axsen, J. and K. S. Kurani (In Press). Interpersonal Influence within Car Buyers’ Social Networks: Applying Five Perspectives to Plug-in Hybrid Vehicle Drivers. Environment and Planning A (doi: 10.1068/a43221). Bartle, C., Avineri, E. and Chatterjee, K. (2011) Information-sharing, community-building and trust: A case study amongst commuter cyclists. In: 43rd Universities Transport Study Group Conference, Milton Keynes, UK, 5th-7th January, 2011. Available from: http://eprints.uwe.ac.uk/16914. Cartwright, E. (2011), Behavioral Economics, 2nd Edition, Abingdon, New York: Routledge. Chorus, C.G. (2012), What about behaviour in travel demand modelling? An overview of recent progress, Transportation Letters: The International Journal of Transportation Research 4, 93-104. Chorus, C.G. and B.G.C.Dellaert (2012), Travel choice inertia Joint role of risk aversion and learning, Journal of Transport Economics and Policy 46(1),139-155. Di, X, Liu, H, Zhu, S, and Levinson, D. (2014) Indifference Bands for Route Switching Presented at Transportation Research Board Annual Meeting, January 2015. Dugundji, E. and Gulyas, L. (2008). “Sociodynamic discrete choice on networks in space: Impacts of agent heterogeneity on emergent outcomes.” Environment & Planning Part B, 35, 1028-1054 Fujii, S. and R. Kitamura (2003). What does a one-month free bus ticket do to habitual drivers? Transportation, 30 (1), pp. 81–95 Fukuda, D. & Morichi, S., 2007. “Incorporating aggregate behavior in an individual’s discrete choice: An application to analyzing illegal bicycle parking behavior,” Transportation Research Part A: Policy and Practice, Elsevier, vol. 41(4), pages 313-325, May Gardner, B., 2009. Modelling motivation and habit in stable travel mode contexts. Transportation Research Part F, 12(1), pp. 68-76 Gifford, J.L. and Checherita-Westphal, C.D. (2008), Boundedly- and Non-Rational Travel Behavior and Transportation Policy, mimeo, George Mason University, Arlington, Virginia, USA. Halpern (2015), Inside the Nudge Unit: How small changes can make a big difference, WH Allen Innocenti, A., P. Lattarulo and M.G. Pazienza (2013), Car stickiness:Heuristics and biases in travel choice, Transport Policy 25, 158-168. E. Mote, Y. Whitestone. The social context of informal commuting: Slugs, strangers and structuration Transportation Research Part A 45 (2011) 258–268 Larrick, R.P. and J.B. Soll (2008), The MPG Illusion, Science 30, 1593-1594. Metcalfe, R. and P. Dolan (2012) Behavioural economics and its implications for transport, Journal of Transport Geography 24, 503–511. Páez, A., Scott, D.M., 2007. Social influence on travel behavior: a simulation example of the decision to telecommute. Environment and Planning A 39, 647–665. Scott, D.M., Dam, I., Páez, A. and Wilton, R.D. (2012) Investigating the effects of social influence on the choice to telework. Environment and Planning A, 44: 1016-1031. Srinivasan, K.K., Mahmassani, H.S., 2000. Modelling inertia and compliance mechanisms in route choice behaviour under real-time information. Transportation Research Record, 1725, pp. 45-53. Steg, L., (2003). Can public transport compete with private cars? IATSSResearch27, 27–35. Sunitiyoso , E. Avineri & K. Chatterjee (2011) On the potential for recognising of social interaction and social learning in modelling travellers’ change of behaviour under uncertainty, Transportmetrica, 7:1, 5-30, DOI: 10.1080/18128600903244776 Thaler, R.H. and C.R. Sunstein (2008), Nudge, Improving Decisions About Health, Wealth, and Happiness, Yale University Press. Wills, T.A., 1981. Downward comparison principles in social psychology. Psychological Bulletin 90 (2), 245–271. Wilson, D. K., Kaplan, R. M., &Schneiderman, L. J. (1987). Framing of decisions and selections of alternatives in health care.Social Behaviour,2, 51–59. Wilton, R.D., Páez, A. and Scott, D. M. (2011) Why do you care what other people think? A qualitative of investigation of social influence and tele-commuting. Transportation Research Part A: Policy and Practice, 45:269-282 Rasouli, S. and H.J.P. Timmermans (2014), Activity-based models of travel demand: promises, progress and prospects, The International Journal of Urban Sciences, 18(1) 31-60.