The Dark Side of Dark Mode -- User behaviour rebound effects and consequences for digital energy consumption

User devices are the largest contributor to media related global emissions. For web content, dark mode has been widely recommended as an energy-saving measure for certain display types. However, the energy savings achieved by dark mode may be undermined by user behaviour. This pilot study investigates the unintended consequences of dark mode adoption, revealing a rebound effect wherein users may increase display brightness when interacting with dark-themed web pages. This behaviour may negate the potential energy savings that dark mode offers. Our findings suggest that the energy efficiency benefits of dark mode are not as straightforward as commonly believed for display energy, and the interplay between content colourscheme and user behaviour must be carefully considered in sustainability guidelines and interventions.

💡 Research Summary

The paper investigates whether the widely promoted “dark mode” actually reduces energy consumption on user devices, or whether user behavior creates a rebound effect that negates any savings. The authors begin by noting that user devices—especially displays—account for the majority of the internet’s carbon footprint, with LCD and OLED technologies behaving very differently: OLED power scales with pixel colour, while LCD power is largely independent of displayed content because of its constant backlight.

To test the hypothesis, the authors conducted controlled laboratory experiments on a 2017 MacBook Pro equipped with a 13.3‑inch LCD. Using a Tektronix PA1000 power monitor, they measured the power draw of the BBC Sounds homepage in both dark and light modes across 16 brightness levels (1 = dim, 16 = full) and repeated each measurement five times. Automated Selenium scripts simulated a page load followed by 15 seconds of scrolling to mimic typical user interaction. The results showed no statistically significant difference in power consumption between dark and light mode at any brightness level (p > 0.05), confirming that on LCDs the colour scheme alone does not affect energy use. Brightness, however, had a strong impact (p < 0.001), with higher brightness leading to higher power draw.

The human‑centred part of the study involved ten participants who were asked to adjust the screen brightness to a comfortable level while viewing the same page in both colour schemes. Tests were performed in two ambient lighting conditions—a dimly lit room and a brightly lit room—yielding four experimental conditions: dark‑dim (DD), dark‑light (DL), light‑dim (LD), and light‑light (LL). Each participant experienced all conditions in a within‑subjects design. Statistical analysis revealed that ambient lighting did not significantly affect the chosen brightness (p = 0.13), but the colour scheme did (p < 0.001). Participants consistently selected higher brightness levels for dark mode (mean ≈ 12.5) than for light mode (mean ≈ 9.6), regardless of the surrounding illumination.

The discussion interprets these findings as evidence of a behavioural rebound effect: users compensate for the darker UI by increasing screen brightness, which on LCD devices directly raises power consumption. Consequently, the presumed energy‑saving benefit of dark mode is nullified, and may even be reversed if users habitually raise brightness. The authors caution that while their results are limited to a single LCD device and a small sample, the phenomenon could be more pronounced on OLED screens where dark mode does reduce pixel power but user‑driven brightness increases could offset those gains.

Limitations include the modest participant pool, the exclusive focus on a laptop LCD, and the short, synthetic interaction scenario (page load and brief scroll). The authors recommend future work that expands the sample size, incorporates mobile devices with OLED displays, examines longer and more varied usage patterns (video playback, gaming), and investigates the role of automatic brightness controls and user perception.

In sum, the study underscores that sustainability guidelines promoting dark mode must account for human behaviour. Energy‑saving claims based solely on technical characteristics of display technology may be overly optimistic if users adapt their settings in ways that increase overall consumption. The paper calls for a more holistic approach that integrates UI design, device hardware, and user experience research to achieve genuine digital carbon reductions.