Is ‘referencing’ a remedy to hypothetical bias in value of time elicitation? Evidence from economic experiments

Abstract

This paper demonstrates that commonly used methods for eliciting value of time can give downward bias and investigates whether this can be reversed by 'referencing' as has been suggested (e.g., by Hensher in Transp Res B 44:735-752, 2010), i.e. with attributes of choice alternatives pivoted around a recently made journey. Value-of-time choice experiments were conducted in two rounds. In the first round, real and hypothetical purchases of performance of a simple time-consuming task were done to assess hypothetical bias; in the second round, participants were asked to do hypothetical travel choices with and without 'referencing' to a specific occasion, to be able to test 'referencing' as a remedy to the bias confirmed in the first round. A negative hypothetical bias was found for allocation of time at another occasion than the present (but not for a decision concern allocation of time 'here and now'). A striking result was held from the second round experiments: 'referencing' indeed affects responses, but by reducing the elicited implicit value of time, so any negative hypothetical bias that would exist without 'referencing' would have been further magnified by the 'referencing' design.

Real consequences matter: Why hypothetical biases in the valuation of time persist even in controlled lab experiments

“…We find that the hypothetical bias identified in our experiment can to a large extent be attributed to participants who have scheduling constraints during the time when the unexpected waiting would take place. This may also explain why Hultkrantz and Savsin (2018) in a very similar experimental setup found no hypothetical bias: they only considered an unexpected waiting time of 15 minutes, in which scheduling constraints supposedly play a much more minor role compared to the range of waiting times considered in this experiment (15-90 minutes).…”

“…Also the underlying data source (SP vs. RP data) seems to affect the value of time significantly (e.g. Ghosh, 2001;Hensher, 2001;Brownstone and Small, 2005;Small et al, 2005;Isacsson, 2007;Peer et al, 2014;Hultkrantz and Savsin, 2018): most studies find that the value elicited from SP data is substantially lower than the value elicited from RP studies. This gap is usually referred to as hypothetical bias.…”

“…We use this controlled environment to test existing hypotheses related to the origin of hypothetical biases. Unlike the recent study by Hultkrantz and Savsin (2018), which has a similar scope, we focus on unexpected waiting times that are long enough to interfere with scheduling constraints (15-90 minutes); in the experiments conducted by Hultkrantz and Savsin (2018), the unexpected time (to be spent on completing a task) was only 15 minutes. We will show that this is an important difference, as scheduling constraints play a crucial role in the formation of hypothetical biases in the valuation of time.…”

“…Regarding the design of SP experiments, attribute values pivoted around the respondents' real-life reference situation have been suggested as a way to reduce hypothetical biases, and have been implemented in numerous applications (Hensher, 2010;Li et al, 2018). However, a recent study by Hultkrantz and Savsin (2018), which also concerns VOT elicitation, provides some evidence that 'referencing' to a specific situation may have little influence on the hypothetical bias, and may actually acerbate it: they find no hypothetical bias to be present if the decision between a 15-minute task and a monetary reward concerns the present moment (which hence assumes the role of a reference point), but indirectly find evidence for a hypothetical bias to exist if the trade-off concerns a later (not precisely defined) point in time.…”

“…They may, for instance, induce drivers to more often choose toll lanes in real life than implied by their SP responses, leading to comparatively lower SP-based VOTs. A further hypothesis is that in hypothetical settings, respondents do not take into account their scheduling restrictions to the same extent they would do in real life (or more generally "disregard alternative time use", see Hultkrantz and Savsin (2018)), due to the lack of real consequences. As a result the SP-based VOT is lower than the RP-based VOT, since the former does not fully account for scheduling constraints.…”