Whimpact: More questions than answers
NB: This article has been prepared in the context of the MaaS National Reference Committee and in Melbourne (May 1-2), coordinated by Intelligent Transport Systems Australia, which will bring together a number of global luminaries including Sampo Hietanen (MaaS Global), Andy Taylor (Cubic Transportation Systems), Colin Lim (SMRT鈥檚 mobilityX), amongst others. The ITLS team are avid contributors to both these forums.
The contents of this article are motivated by the report , prepared by Ramboll, which documents behavioural impacts from MaaS Global鈥檚 Whim application in Helsinki, Finland for the period January-December 2018.
Whimpact has been clogging up my LinkedIn feed. Everyday, I am seeing industry figures citing the report and holding it high on a pedestal in ways as though it were gospel. There exists a misguided notion that the report, whilst valuable, serves up a on the many potential follies of 鈥攖hat it may hurt public transport patronage, lead to class segregation (by targeting those living in the urban core), increase single-occupancy vehicles as well as the number of trips people make鈥攁ll with associated externalities.
The gist of the report (and what have been oft-cited) are the number of conjectures offered up. We present our own observations based on these assertions.
| # | Assertion | Observation |
|---|---|---|
| 1 | MaaS users ride public transportation more than their Helsinki counterparts | Perhaps public transport users are more likely to subscribe to MaaS? |
| 2 | MaaS users are multimodalists | Perhaps multimodalists are more likely to subscribe to MaaS? |
| 3 | MaaS helps solve the first/last mile problem | What modes did these subscribers use previously? |
| 4 | Taxis are a welcome option to MaaS users | Could taxi users be more likely to subscribe? In the unlimited model, what entices people not to choose taxi? |
| 5 | MaaS users make shorter citybike trips | What role might visitors play? |
| 6 | Average daily trips of MaaS users and typical Helsinki residents are about the same | This is an average but how about the variance in daily trips? |
| 7 | Public transportation is the backbone of MaaS | This is the goal but where is the evidence? Is this 鈥榖ackbone鈥� designed into the MaaS product initially? |
| 8 | MaaS grows along public transport corridors | Is this another case of self-selection bias? |
| 9 | New transportation solutions can replace 38% of daily car trips | This is an estimate based on behavioural assumptions |
| 10 | MaaS users play by the rules | A funny way to put it. How to detect fraud and enforce compliance is a big issue |
| 11 | Rental cars are part of MaaS daily trips | Why are rentals more attractive than carsharing? |
The conclusions drawn in the Whimpact report suffer from a number of shortcomings. Firstly, there is the notion that MaaS causes behavioural change. This is a massive logic issue in terms of causality since it is likely the case that people of a particular behavioural paradigm (i.e., existing public transport users) self-select to become subscribers of MaaS. This links to the second issue around the inability to define the counter-factual which makes any sorts of valid arguments difficult around the potential impacts of MaaS on travel behaviour.
Thirdly, there are not enough plan options being tested. There are three tiers of service offered in Helsinki: Whim to Go as a pay-as-you-go service with no monthly fee; Whim Urban (鈧�49 per month) with unlimited public transport, max 鈧�10 for taxi (within 5 km), fixed daily 鈧�49 rental car fee and free city bikes (30 min max); and Whim Unlimited (鈧�499 per month), featuring unlimited public transport, taxi (within 5 km), rental cars and city bikes (30 min max). Note that technically, only the latter two constitute 鈥榮ubscription鈥� models in the pure sense of the word.
Under Assertion 6, the report makes the point 鈥渢here has been some speculation that unlimited MaaS packages might lead to a major uptick in total trips鈥攑articularly taxi鈥�. This is one of the greatest risks in deploying MaaS as the provider nudges consumers onto more expensive and lucrative modes. Who wouldn鈥檛 default to taxi (or an equivalent point-to-point option like Uber) as their first port-of-call if they had already purchased an unlimited transport bundle?1聽The report goes on to state that the 鈥渄ata suggest this is not the case鈥�, but buried deep in the Appendix it turns out that Whim Unlimited users were too few and so excluded from the analysis proper.
This links to the final question of what exactly the sample size was for this Helsinki study. The report points to 鈥渙ver 70 000 registered users鈥� of Whim. How many of these are pay-as-you-go (PAYG) customers? How many are regular users? We suspect a large number could constitute app downloads only since the report leaves this undefined.2 The report offers the below figure (page 53) showing a relative increase in uptake but offers no scale on the vertical axis.
The Helsinki experience could offer immense insight into the two greatest points of contention in the MaaS debate. The first relates to the competing merits of subscription versus PAYG models. We are not sure what to infer from Whimpact but in other markets (e.g., West Midlands) Whim only offers a PAYG option. Vij et al. (2018) found that PAYG is more attractive but in our own ITLS research we saw that subscription models were preferred (Ho et al., 2018)鈥攊n fact, up to 47% of our sample will subscribe. Consumer preference for these alternatives will likely depend on the design of subscription plans and the discounts built in as part of the MaaS offering.
This is an interesting question since it raises the issue of business models as the original concept is for a MaaS business to purchase transport assets/capacity in bulk and thus negotiate to obtain quantity discounts. They are able to finance this with customers pre-paying for their mobility credits, and based on the assumption that some of these credits will remain unused at the end of each period (think phone plans), all whilst passing on a small proportion of the original discount to the consumer (as occurs with Whim Urban). If PAYG is indeed the preferred approach, then what is the value add for the customer beyond a better journey planner? Will this model even be sustainable in the long run?
Linked to business models is the second point of contention around whether government or the private sector ought to take the lead on MaaS. There are many schools of thought on this question, but we think that a government model will lead to issues of bias and stifle innovation (UITP takes another view, of course). Whether government even wishes to be involved in the service provision space is another question鈥攚e have explored these questions in detail for over 30 years through the . Our latest idea is what we term a MaaS broker/aggregator which brings together specialised businesses including transport operators, technology providers and financial enterprise. This is very much a private sector-driven model and indeed we find that private enterprise prefers government involvement at arm鈥檚 length only鈥攕trategic support legitimising the sector and ensuring a fair and level playing field, as opposed to any specific monetary support like subsidies from government (Wong et al., 2019).
At ITLS, we are launching an exciting , jointly developed with IAG and SkedGo through the iMOVE Cooperative Research Centre (CRC). Unlike other (often commercially-driven) pilots and trials around the world, the design of our trial has a real emphasis on research and in garnering robust behavioural insights. We will bring in the most sophisticated analytical capabilities to tackle the many unknowns plaguing the sector. What are the barriers to forming MaaS businesses? How attractive are subscription models versus PAYG? With proper before and after data collection, we avoid the issues of self-selection bias plaguing Whimpact and can finally ascertain empirically whether there is any evidence that MaaS can attract car users onto public transport. Our findings will help determine whether we can dismiss the potential follies of MaaS, or might we need extensive government intervention to ensure that MaaS development occurs in a way consistent with common societal objectives (Wong et al., 2017).
We are looking forward to some very fruitful conversations during the week.
Acknowledgements: I thank Christoffer Weckstr枚m (Aalto University) for earlier discussions on this topic and Corinne Mulley (ITLS 鈥楨ngland鈥�) for additional comments and advice.
[1] We note heavy congestion might make taxi less attractive than rail or rapid bus.
[2]聽There are 50K+ downloads of the Whim application on the Google Play store.
References and resources
HO, C., HENSHER, D. A., MULLEY, C. & WONG, Y. Z. 2018. Potential uptake and willingness-to-pay for mobility as a service (MaaS): A stated choice study. Transportation Research Part A: Policy and Practice, 117, 302-218.
RAMBOLL 2019. Whimpact: Insights from the world's first mobility-as-a-service (MaaS) system. Copenhagen, Denmark.
VIJ, A., RYAN, S., SAMPSON, S. & HARRIS, S. 2018. Consumer preferences for Mobility-as-a-Service (MaaS) in Australia. 40th Australasian Transport Research Forum (ATRF). Darwin, Australia.
WONG, Y. Z., HENSHER, D. A. & MULLEY, C. 2017. Emerging transport technologies and the modal efficiency framework: A case for mobility as a service (MaaS). 15th International Conference on Competition and Ownership in Land Passenger Transport (Thredbo 15). Stockholm, Sweden.
WONG, Y. Z., HENSHER, D. A. & MULLEY, C. 2019. Mode-agnostic mobility contracts: Identifying broker/aggregator models for delivering mobility as a service (MaaS). 98th Transportation Research Board (TRB) Annual Meeting. Washington, D.C., United States.