Is information ignorance a strength?

In today's service industries, realtime delay information is ubiquitous. The waiting time to cross the US and Canada border is posted online and updated in real-time. Information about traffic jams on major roads is distributed on radio, television, and the internet. Thanks to trafficinformation- sharing apps like Google Maps and Waze, real-time traffic information is even available for roads that are not covered by government-funded traffic detection and monitoring.

Not everyone uses the information

Regardless of how widely real-time delay information may be available, a large percentage of customers remain uninformed. This may be due to information ignorance or rational behaviour in the presence of an information cost. For example, people may overlook up-to-theminute information about delays before setting out. In an online poll, some 20,000 participants were asked, "How often do you check traffic information before going out?" 47% answered, "I don't check"; the rest checked various sources, such as TV, radio, computer, or mobile devices . Some people may be over-confident that they will be lucky. Some may check for information now and then. Another reason for information heterogeneity could be that small service providers may not be able to afford the technology for tracking and reporting real-time delay information. In that case, only dropin customers can see the queue, whereas many potential customers cannot. Lastly, a substantial minority still do not possess a mobile device.

So, customers possess varying degrees of knowledge about realtime delays: some are fully informed, others completely out of the picture. In order to understand the impact of delay information on society, it is essential to investigate the interaction among customers in a system characterised by information heterogeneity.

Conventional wisdom

Conventional wisdom suggests that delay information benefits society. The intuition is that congestion information helps better match capacity with demand: customers tend to avoid highly congested roads and follow more free-flowing routes. That rationale is consistent with the prevalence of real-time delay information in today's public service industries. However, we argue that it is unrealistic to expect that all customers have access to real-time delay information even if it is readily available. More importantly, does the system necessarily suffer efficiency-loss from the presence of uninformed customers?

Reassessment of conventional wisdom

We investigate a queueing model with two streams of customers who differ in their information structures in order to answer these questions. Informed customers decide to join or balk based on real-time delays. Uninformed customers are not aware of the real-time information and base their join-or-balk decision on their experience of long-run average delay. We characterise customers' equilibrium behaviour with information heterogeneity and investigate how the presence of a larger fraction of informed customers affects the overall social welfare of the system.

Does real-time delay information help society? Think again

Contrary to the conventional wisdom that real-time congestion information always improves social welfare, we discover that social welfare is "unimodal," that is, it first increases and then decreases in the information level when the system experiences a high enough offered load. When the offered load is relatively low, the conventional wisdom is true – information prevalence always benefits social welfare. This is because growing information prevalence has both positive and negative effects on social welfare. On the positive side, if the real-time system congestion is visible to customers, system capacity can be more efficiently matched with customer demand intertemporally because informed customers seek service only if the service benefit dominates the waiting cost – when the queue is short. However, informed customers' self-interested joining behaviour might overload the system, especially when the customer arrivals are overwhelming. In this situation, uninformed individuals' presence mitigates system congestion: these uninformed customers are reluctant to join a busy system without realtime information. This disincentive helps free up the capacity to serve more informed customers, who contribute more surplus to the overall efficient running of the system.

Nonetheless, when a large proportion of a high customer volume becomes informed, uninformed individuals eventually lose interest in the service. As information prevalence grows, the system suffers from rising externality inflicted by an increasing fraction of informed individuals if they all choose to balk. Hence the system's overall functioning deteriorates because of growing information prevalence.

Our results highlight the fact that some degree of real-time information heterogeneity in the population can lead to more efficient outcomes in terms of social welfare than can information homogeneity. The presence of uninformed customers does not necessarily harm the system. It increases social welfare when the system experiences high offered loads. Our results also imply that there may be value in intentionally introducing information heterogeneity and controlling realtime delay information availability.