Little is known about how effort is represented for different kinds of tasks. Recently, we suggested that it would help to establish empirical benchmarks for this problem. Accordingly, Feghhi and Rosenbaum (Journal of Experimental Psychology: Human Perception and Performance, 45:983–994, 2019) estimated how many additional digits to be memorized corresponded to navigating through a narrow gap versus a wide gap. The estimates were based on a study in which participants chose between walking paths with associated memory demands. We found that participants were equally willing to choose to walk through a narrow gap as to walk through a wide gap when the narrow-gap walk required memorization of 0.55 fewer digits on average than the wide-gap walk. In the present experiment, we sought to replicate and extend this previous finding in two ways: (1) by presenting the memory digits in auditory rather than visual form to test the hypothesis that participants used phonological recoding of the visually presented digits; and (2) by providing a new metric of the relative difficulty of navigation errors compared to recall errors. We provided 36 university students with two action/memorization options per trial and asked them to choose the easier option. Each option had varying degrees of physical demand (walking through a wide or narrow gap) and mental demand (memorizing 6, 7, or 8 digits). We expected performance to be comparable to what we observed earlier with visually presented digits to be memorized, and this prediction was confirmed. We also used a new metric to show that navigation errors were implicitly judged to be 17% more costly than recall errors. The fact that this percentage was not 0 indicates that reducing percent error was not the only basis for reducing effort.