BEHAVIOURAL RESPONSES TO THE INTRODUCTION OF A NEW TRAVELCARD IN DUTCH PUBLIC TRANSPORT: THE CASE OF AMSTERDAM

Size: px
Start display at page:

Download "BEHAVIOURAL RESPONSES TO THE INTRODUCTION OF A NEW TRAVELCARD IN DUTCH PUBLIC TRANSPORT: THE CASE OF AMSTERDAM"

Transcription

1 BEHAVIOURAL RESPONSES TO THE INTRODUCTION OF A NEW TRAVELCARD IN DUTCH PUBLIC TRANSPORT: THE CASE OF AMSTERDAM Dr. Barry Ubbels NEA Transport research and training, the Netherlands Drs. Nathaly Dasburg-Tromp NEA Transport research and training, the Netherlands 1. INTRODUCTION The Netherlands has a public transport system comprising bus, metro, tram, train and a few ferry services. In the past, travellers had the choice between a wide range of paper tickets with different fare levels. The fare price was amongst others dependent on distance travelled (the number of zones (regional public transport) or the number of train kilometres), frequency of travel (e.g. monthly travel passes), the purchase location, age of the traveller and time of day (railways). There was no fare integration between regional public transport and national railways: the well-known Strippenkaart was and is generally not valid in the train. About a decade ago, the Dutch government together with the transport operators decided that there was a need for a new travelcard system supported by new chipcard technology. This new travelcard should support seamless travel using multiple public modes and be easy to use by travellers. This card is called the OVchipkaart (translated: public transport chipcard) and is a joint initiative of the public transport operators. It has been gradually introduced in the Netherlands starting in 2005 with the region of Rotterdam. More regions followed where both old paper tickets as well as the OV-chipkaart were accepted. As of june 2010, it is not possible to use paper tickets in the regions of Amsterdam and Rotterdam. It is expected that the Strippenkaart will not be valid in the Netherlands by mid Besides fare integration, the OV-chipkaart also introduced a new fare structure. Before introduction, regional public transport fares were based on the number of zones travelled, whereas with the OV-chipkaart people pay according to the number of kilometres travelled. Introducing a new smartcard with a new fare structure may have various consequences. Public transport demand may change because of higher or lower (perceived) prices, or travelers dislike the use of new technology and decide to change mode. This may result in a change in revenues for the operator and administration costs. It may well support faster boarding times or reduce fraud levels. In this study we are interested in the behavioural consequences of the introduction of the OV-chipkaart. There are two major differences with the previous situation which may change travel behaviour. On the one hand there is the introduction of smart payment (which involves a total new way of payment and activities: loading of a sufficient travel budget on the OV-chipkaart, checking in and checking out in vehicles, etc.), on the other hand the introduction coincides with new fare levels (according to the number of kilometers travelled). There are not many examples in the world where a smartcard payment system is introduced nationwide, which integrates fares and changes the fare structure at the same time. We are aware of the Oyster card in London and the Octopus card in Hong Kong, but as far as we know, these have been introduced with the only aim to

2 integrate ticketing. The effects of fare integration have been analysed in a literature review of case studies from major urban areas across Europe, North America and Australia that have qualified and quantified the benefits of simple, affordable and competitive integrated ticketing schemes implemented between the early 1980 s to 2008 (Booz & company, 2009). This overview provides evidence that support increased patronage from integrated ticketing in public transport. Simplified and integrating ticketing lead to increased levels of travel demand in the range of 6% to 20%. This demand seems not to results from a modal shift effect: there is only limited quantitative evidence to support a link between modal shift and fare integration. While fare integration has behavioural effects, there is also ample evidence that fare changes cause changes in demand for travel. Price sensitivity in public transport has been addressed in literature. Economists often report price elasticities of public transport demand. PBL and CE (2010) give an useful overview of Dutch and international elasticities. It appears that elasticities are in the range between -0,2 to - 0,5 in the short run (with little difference between Dutch and international cases). This means that a price increase of 10% leads to 2% to 5% less demand. There is a difference between modes, with metro and tram demand being less price sensitive than the demand for bus and train. This paper describes the consequences of the OV-chipkaart introduction in the Netherlands, and more specifically in Amsterdam. As mentioned before, this new card is currently introduced in phases in the Netherlands. After a period with both paper tickets and OV-chipkaart payments accepted, it is now only possible to pay with the OV-chipkaart for regional public transport (bus, tram and metro) in the greater Amsterdam region. In 2009, it was the Amsterdam metro that became inaccessible without OV-chipkaart as a first step in a process towards overall implementation in this region. As of the 27th of August 2009 it was only possible to travel with the Amsterdam metro when using the new OV-chipkaart. This was a very interesting moment to analyse travel behaviour because (realistic) public transport alternatives without chipkaart obligation were available (bus, tram and train). The aim of this paper is to analyse behavioural responses and price perceptions of the public transport user in the greater Amsterdam region. Implementation of the chipkaart changes fare levels, and we anticipate that perceptions of the changes (which might be wrong) are important drivers for travel decisions in practice. This paper presents the qualitative results form a survey amongst public transport users in Amsterdam. We have compared travel behaviour before the obligation to use the chipcard in the Amsterdam metro (ex-ante) with travel patterns after the 27 th of August (ex-post). This paper is organized as follows. Section 2 explains the survey approach and discusses the results of the ex ante measurement. Section 3 presents the results from the second survey (approaching the same group of travellers) which started a few weeks after the 27 th of August. Section 4 concludes and gives some policy recommendations and issues for further research. 2. Expectations: travel behaviour and fare levels 2.1 Public transport in Amsterdam Amsterdam has a well developed public transportation network. Travellers may decide to use bus, tram, metro, train and/or ferry to reach their destination. Within the metropolitan region there is much commuting from outside Amsterdam into the city. Public transport within Amsterdam is operated by the GVB (owned by the

3 municipality), which provides bus, tram, metro and ferry services. The public authority responsible for public transport in the greater region is the city region of Amsterdam (Stadsregio Amsterdam). This is a decentral governmental organisation, responsible for public transport in the region. It tenders concessions which allows firms to periodically operate public transport in certain regions (outside Amsterdam). There is much commuting from these regions to the city of Amsterdam which makes these concessions attractive. Fare levels in the Netherlands have long been regulated by the national government, but the introduction of the chipkaart comes with a tendency towards decentralisation of fare regulation. Demand for public transport in the Netherlands is highest in the Randstad area, most people live there and public transport availability is high. Amsterdam has about citizens (2010), and the Metropolitan area of Amsterdam about 1.4 million (Stadsregio Amsterdam, 16 municipalities). It is one of the heaviest congested areas of the Netherland, which makes public transport a realistic alternative. Compared with other regions, most public transport with bus, tram and metro takes places in Amsterdam. In 2008 more than 1 billion traveller kilometres were registered for the city region, more than twice as much as the region of the Hague (NEA, 2010). In 2009 on average about trips per day were made by travellers using the metro in Amsterdam. Tram use is higher with trips per day. 2.2 Data collection and survey: the OV-panel This study focuses on the consequences of the mandatory use of the OV-chipkaart in the Amsterdam metro. This may have positive effects: less check in times and increased safety levels (access to the platform was only possible when a OVchipkaart with sufficient travel budget opened the gates) may attract more users to the metro. On the other hand, travellers may have some practical difficulties (one needs to check in and check out), and they might be afraid that prices increase. This is an incentive for metro travellers to shift to other modes (bus and tram use was still possible without chipkaart) 1 or prefer slower modes (bicycling is an attractive alternative in Holland, especially in summer when the weather is better). We will come back to the change in fare structure later. Governments have promised that on average trips will not become more expensive with the OV-chipkaart. However, since zones were not equal in distance, it is clear that large zones are more expensive than small zones in a new situation with kilometre prices. Also detouring of buses and trams becomes relatively more costly to travellers. For the purposes of this paper we are interested in the behaviour of those who have to change from a paper ticket to an OV-chipkaart: the metro users in Amsterdam. NEA owns a panel of almost 40,000 respondents (at the time of the survey, about 32,000 respondents) which is representative for the Dutch public transport traveller who makes use of bus, tram and metro (train users are not included in the panel). Only students (using the OV-studentcard) seem to be slightly underrepresented in the panel. Since we know in which region the respondents make journeys, it is possible to make a selection of public transport travellers in the greater Amsterdam region who are potential Amsterdam metro users. About 5200 respondents in the panel appeared to travel in Amsterdam and received an invitation to participate in our Internet survey. Two surveys have been carried out: a few weeks before and after the 27 th of August. The first survey contained questions about ticket use, and existing and expected 1 As of June 2010 it is not possible anymore to use public transport at all without OV-chipcard in the city region of Amsterdam.

4 travel behaviour. The second survey asked for actual travel behaviour after chipkaart obligation in the metro. Since we do know most relevant socio-economic characteristics of all panel members (such as age and education) there was no need to ask for this data. The response was reasonable. The first survey was responded by 2,969 travellers (a response of 57%) of which 1508 often used the metro for their most common journey. The second survey was directed to those travellers of which we knew from the first survey that they did use the Amsterdam metro (more or less frequently). This selection resulted in 2,551 travellers, of which 1,840 responded (72%). This chapter continues with the most interesting results from the ex ante survey, the next chapter presents the ex post survey results and compares. For more detailed results, we refer to the full NEA report (Scheltes et al., 2009). 2.3 Ex-ante survey: perceptions and travel intentions We start with some relevant statistics of the respondents from the first survey: the Amsterdam public transport traveller. It appears that the educational level is somewhat higher than for the average Dutch traveller. About 64% of the respondents uses public transport for commuting reasons. Shopping and social purposes are other frequently mentioned reasons to travel. Most respondents use multiple travel modes for their journey. This can be explained by the fact that there is much commuting from the region into Amsterdam. In many cases train and bus are used to travel to, for instance, Amsterdam Central Station, where one changes to tram or metro to reach the final destination. Relatively more elderly have responded, 70% is older than 41. The ticket that is being used to travel is also relevant. It was already possible to travel with the OV-chipkaart (most travel cards can be uploaded on the OV-chipkaart, also travel cards for a month or a year). About 26% of the respondents used the Strippenkaart. Most respondents (at least 44%) have a travel pass which gives them unlimited use of public transport for a certain period of time (year, month or week). This corresponds with the relatively large number of commuters in the sample. The OV-chipkaart is already used by 17% of the respondents. Metro users seem to travel more with the chipkaart than travellers who do not use the metro (20% vs. 14%). This seems an effect from efforts of the Amsterdam public transport operator to make travellers aware of the upcoming obligation to use the chipkaart and to persuade them to change already in an early stage Expected travel behaviour About two weeks before the obligation to use the OV-chipkaart in the Amsterdam metro, we have asked the Amsterdam public transport user if they will adjust their travel behaviour because of this obligation. Table 2.1 shows the percentages. Most of the respondents indicate that they will not change their travel pattern. However, about 15% of all travellers, and 13% of the frequent metro users, indicate to travel less by metro when they must use the OV-chipkaart. Only a very few people expect to make more use of the metro.

5 Table 2.1: Expectations of travellers to adjust metro use Overall Frequent metro user No adjustment 70% 84% More metro use 1% 1% Less metro use 15% 13% Never by metro 12% -- No answer, do not know 2% 2% Total 100% 100% Number of respondents It is also relevant to know (e.g. for the GVB) about the reasons behind this choice, the number of trips that will be adjusted (frequent travellers or not) and what alternatives will be chosen. We only consider the group who indicates to travel less. A majority of about 40% indicates that the trip has become more expensive with the OV-chipkaart. This is confirmed by the relationship between price expectation and expected travel behaviour: almost all respondents who indicate to travel less think that fare prices will increase. Other reasons for not using the metro anymore include the (low) user friendliness, the unreliability and privacy issues. Finally, it seems that about 3 metro trips per week will be adjusted and that most travellers will make use of the tram (22%) and the bicycle (20%) instead of the metro. About 51% of the respondents indicate not to travel by metro, bus or tram anymore, which implicitly means that this group will not travel anymore with the GVB. This demand may be lost due to the OV-chipkaart introduction Fare prices: expectations We have seen that the ticket price is an important reason to make less use of the metro, travellers expect the trip to be more expensive when using the OV-chipkaart. At least they think so, but is this actually true? In order to analyse this we have asked the respondents about the details of their most common metro-trip. This provided information about the station where the respondent entered and exited the metro. This origin-destination data enables to calculate trip distances, we find an average metro trip distance of about 6 kilometre. Since we know the type of ticket that travellers use in the existing situation (before the 27 th of August), it is possible to compare fare levels for at least the Strippenkaart and the new OV-chipkaart 2. Fare levels will change due to the OV-chipkaart introduction. This is caused by the introduction of a new fare structure and moving from a zonal system (Strippenkaart) to a kilometre based structure. First we have asked (all) the respondents whether they think that prices for their trip will change (and if so, whether it would be cheaper or more expensive). About 52% expects OV-chipkaart prices to be higher than existing price levels. Only 8% is positive and thinks that prices will be lower, 25% indicates to expect no change. The correctness of these expectations have been tested for Strippenkaart users in the Amsterdam metro based on their current travel behaviour (355 out of 1508 metro users). 2 The 15 stripes blue Strippenkaart was 7,30 at that time; the fare level of the OV-chipkaart in the Amsterdam region was 0,75 (fixed per trip) plus 0,10 per kilometre.

6 Table 2.2 : Fare expectations and average actual fares for the Strippenkaart traveller Average trip costs Expectation Strippenkaart OV-chipkaart Difference No difference 1,50 1,41-0,09 Lower with chipkaart 1,43 1,49 + 0,06 Higher with chipkaart 1,47 1,44-0,03 Do not know 1,54 1,52-0,02 Average tripcosts 1,48 1,45 + 0,03 Table 2.2 just shows that price differences, as an average for the various groups, are rather small. Note that price differences are caused by the characteristics of the old zonal system and the distances travelled. Long distance trips tend to be more expansive. Since most metro trips tend to be rather short, prices are in most cases not higher but lower which is in contrast with the expectations for most people. Table 2.2 gives results on an average level, but it may be even more interesting to analyse the differences on an individual level. Table 2.3 presents percentages of travellers and compares expectations with the fare implications given their travel behaviour with the Amsterdam metro. These results confirm the previous finding that most respondents have incorrect price expectations and that prices will change (for only 2% of the 300 Strippenkaart users there is no difference in price). About 63% of the (largest) group who expects a price increase will in practice be confronted with a price decrease when using the OV-chipkaart. This also holds for 73% of the group who thinks that prices will not change. In total, about 65% is cheaper off with the new OV-chipkaart. Table 2.3: Estimate of fare developments and actual fare prices for the Strippenkaart traveller OV-chipkaart fare is actually: Absolute number Expectation Cheaper More Equal expensive No difference 73% 25% 2% 93 Lower with chipkaart 53% 47% Higher with chipkaart 63% 35% 2% 178 Total (absolute numbers) OV-chipkaart only: actual use of the Amsterdam metro One month after the 27 th of Augustus, when travel patterns should have adjusted to the new situation, we returned to the Amsterdam public transport user with a new survey to analyse whether forecasted behaviour is in line with metro use in reality. Those travellers who indicated in the ex-ante survey to make less, more or equal use of the metro have been invited to responded to this second questionnaire. Respondents who never used the metro have not been included. About 72% of the 2551 public transport users invited responded which results in a response of 1840 people. The response from those who expected to travel more (in absolute and relative terms) was lower than the other groups of expectations (see Table 3.1). Still, the most interesting group (in terms of changing travel behaviour: those who indicate to travel less) is also here represented with a sufficient number of respondents.

7 Table 3.1: Number of respondents compared Ex-ante expectations Ex-ante survey Ex-post survey Share No change ,9% Less ,5% More ,3% Total ,1% Travel behaviour in reality Most respondents expected not to change their travel behaviour when the use of the OV-chipkaart would be obligatory (see Table 2.1). Table 3.2 shows that 83% of the group with this expectation did, indeed, not change their metro use in reality. The remaining 17% of this group did change their behaviour and decided to travel less or more by metro (which is in contrast with their expectation). In total, about 74% of all ex post respondents decided not to change their metro use. This also includes travellers who expected to change their behaviour before, but in reality they did not. This number is smaller than expected based on the sample (the ex-post survey sample consisted of 82% no changers ). Note that the figures in table 3.2 are answers to the general question whether people changed their behaviour. This is not necessarily caused by the introduction of the OV-chipcard (this will be analysed below). The increase in number of people that do change their metro use may be explained by the fact that people have other reasons for avoiding the metro than the chipkaart obligation only. Interestingly, about 30% of the group who expected to make less use of the metro before, did not change behaviour in practice. More than 68% of the people who indicated to travel less in the ex-ante survey, actually travelled less or not at all after chipkaart introduction. Table 3.2: Metro use after OV-chipkaart obligation (not specifically due to OVchipkaart) Ex-ante: no Ex-ante: less Ex-ante: more Total Use of metro after change use use 27 th August Number % Number % Number % Number % No change More metro use Less metro use No use at all Total When considering all ex-post respondents, it appears that almost 25% decided to travel less or not at all. However, note that the figures in table 3.2 are answers to the general question whether people changed their behaviour. This is not necessarily an effect of the obligation to use the OV-chipkaart, people may have other reasons not to use the Amsterdam metro system. This becomes clear when asking for people s motivation. Still, about half (54%) of those who decided to actually make less use of the metro indicates that this is caused by the OV-chipkaart. In other words, about 13% of all the surveyed travellers (243 out of 1840) decides to travel less by metro explicitly due to the introduction of this card. This is less than what was expected from the ex ante survey (315 respondents expected to travel less). Most respondents indicated that the OV-chipkaart is expensive (in 31% of the cases) and difficult to use (20%). At least people think that it has become more expensive to use public

8 transport, but we have seen before that this is not necessarily true. It seems that perceptions may drive behaviour (we come back to this in 2.3.2). What does this change in travel behaviour mean for the demand of metro travel (in number of trips)? It makes a difference when frequent travellers decide to change. In most cases, the change in the number of trips per traveller is limited to one or two a week. Nevertheless, when we relate the total number of trips that is adjusted to an approximation of the total number of trips made by the sample, and not considering other effects (e.g. implementation of the OV-chipkaart in the metro presumably leads to more people paying for their trip), we may roughly assess the effect of OVchipkaart introduction at about 5% (less demand in number of trips). This is considerable to any public transport operator. Note that not all metro trips have been replaced by car trips. Bus and tram are popular alternatives for 50% of the trips, and these modes of transport are operated by the same company in Amsterdam. The car is chosen as alternative in 13% of the cases, walking and bicycling have a share of 30% Fare prices after obligation The previous chapter has shown what the expectations are: before the chipkaart obligation most travelers thought that fare levels would increase. This perception affected the expected travel behaviour. The difference between the ex-ante and the ex-post situation is that now every metro user needs to use the OV-chipkaart and travelers should at least be able to retrieve what the amount is that they pay for their metro trip. If they want, fare levels are known. It is now interesting to compare expectations with the answers to the question how fare levels have changed (after the 27 th of August) 3. Table 3.3: Ex-ante and ex-post answer to fare level development of metro user (absolute numbers) Actual OV-chipkaart fare Equal Cheaper More Do not Total Expectation (ex-ante) expensive know No difference Lower with chipkaart Higher with chipkaart Do not know Total A set of 1055 metro users have answered both sets of questionnaires, which allows comparison. Table 3.3 shows what these respondents expected before, and what their answer is in the ex-post survey when they (could) have fare information from actual usage. The diagonal numbers are those respondents whose expectation is equal to their answer about the actual fare price they pay. It is very interesting to see that travelers do have their expectations, but when it comes to reality they actually do not know what they pay (485 after vs. 159 before). About 45% is unsure about the fare level they pay for their metro trip. This corresponds with earlier research that finds that OV-chipkaart fare levels are difficult to understand and non-transparent. It seems that a majority has moved from a more expansive expectation to uncertainty 3 Note: it is still possible that respondents have repliedd based on expectations or perceptions, and not on factual knowledge about fare levels.

9 now that the card is there. The share of pessimistic travelers has dropped from about 50% to 20%. 4. CONCLUSIONS A new public transport ticketing scheme, supported by chipcard technology, is stepswise introduced in the Netherlands to replace the traditional paper ticket payment (strippenkaart). Travellers are confronted with a new card and a new fare system where price depends on the kilometres travelled. A few years ago Amsterdam introduced the OV-chipkaart as a valid ticket in addition to the old paper ticket system. A first step in the complete outroll of the OV-chipkaart payment was the obligation to use the chipcard in Amsterdam metro as of the 27 th of August 2009 (no paper tickets accepted anymore). This paper reports about the behavioural intentions and actual responses of public transport users in Amsterdam to the OV-chipkaart obligation in the Amsterdam metro from a before and after survey. Specific attention has been given to the perception of price changes and the impact for travel behaviour. The ex-ante study reveals that most travellers (about 70%) expect not to change metro use due to obligation of chipcard payment in the Amsterdam metro. Still, about 15% expects to travel less by metro. Reasons for travelling less include increased costs (in 40% of the cases), difficult to use and unreliability of the new card. This cost element is interesting, because we find evidence for a relationship between travel behaviour and price expectations: those who expect to travel less also expect a price increase. But these price expectations appear in many cases to be incorrect. For a group of Strippenkaart travellers we found that 63% of those who expect a price increase will in practice be cheaper off. A wrong price perception drives travel behaviour. The ex-post study shows that not everybody who expected to travel less did so in practice. Nevertheless, about 13% of the public transport travellers has travelled less by metro due to the obligation to use the OV-chipkaart. Again costs and userfriendliness are mentioned as most important reasons. Bus and tram have been popular alternatives (operated by the same company as the metro and at that time accessible with paper tickets), as well as walking and cycling. Only 13% changes to car transport. Note that the new chipcard also has positive effects for revenues of the operator because it is more difficult to travel without a valid ticket. Finally, we asked about (actual and not expected) fare prices now that everybody has to use the OVchipkaart. It appears that most travellers are uncertain about fare prices. Most have moved from a more expensive expectation to uncertainty in practice. Note, however, that travellers may also be uncertain about fare levels in the existing situation with paper payment. We find that many respondents have a wrong perception of fare prices in the new situation (given individual travel patterns), which seems to affect travel intentions and actual travel behaviour. Also after introduction there is much uncertainty about fare changes. The Dutch government claims that on average the traveller will not be worse off with the new fare system. Apparently, the traveller is not convinced and communication about fare structure and fare levels needs to be improved because it may lead to a reduction in travellers using public transport. This paper has discussed the qualitative results from a rich data set. Further research will focus on statistical evidence of the relationships that have been identified and other explanatory research (is education important for price perception and travel

10 behaviour?). It would also be interesting to analyse the introduction of the OVchipkaart in other parts of the Netherlands: what are price perceptions and travel behaviour. An important issue for governments is communication: will this change perceptions into the correct direction? References Booz & co. (2009) The benefits of simplified and integrated ticketing in public transport, report prepared for the public transport executive group, draft report, United Kingdom. NEA (2010) Ontwikkeling openbaar vervoer: vraag en aanbod, Zoetermeer (in Dutch). PBL (Planbureau voor de Leefomgeving) and CE Delft (2010) Effecten van prijsbeleid in verkeer en vervoer, een kennisoverzicht, Den Haag (in Dutch). Scheltes, W.H., B.J. Ubbels, & N. Dasburg-Tromp (2009) De Amsterdamse metro; van strippen naar chippen: effecten van de verplichte invoering van de OV-chipkaart, NEA Transport research and training, the Netherlands (in Dutch, not public).