energy · environment · family: the rulebook was written for a world that no longer exists
brighter bulbs, more light, same problem
LED efficiency gains backfire as total lighting energy and light pollution surge
Problem statement
LED technology achieved a ~7× improvement in luminous efficacy over incandescent bulbs (15 lm/W to 100+ lm/W), reducing the cost of producing one lumen-hour by 85–90%. But this success triggered a massive expansion in total light consumption rather than a proportional reduction in energy use. Over three centuries and five lighting technologies, humanity has consistently spent approximately 0.72% of GDP on lighting — every efficiency gain has been absorbed by increased consumption (the Jevons paradox or "100% rebound"). LED-era evidence confirms the pattern is continuing, with the additional harm that cheap, blue-rich white LEDs may be driving ecologically damaging light pollution at rates far exceeding what satellite monitoring can detect.
Why this matters
Night-sky brightness is increasing at 9.6% per year (equivalent to doubling every 8 years), based on 51,351 citizen-scientist observations (2011–2022). This is far worse than the ~2% per year measured by satellites, in part because satellites miss the blue-light component of LEDs. A 4000K white LED is approximately 2.5× more ecologically disruptive for scotopic (night-adapted) vision than high-pressure sodium lighting. Despite continued efficacy improvements, global lighting energy consumption increased in 2022, particularly in large emerging economies (IEA). The rebound eliminates the climate benefit of LED adoption while creating a new category of ecological harm.
What’s been tried and why it hasn’t worked
Energy efficiency standards (e.g., EISA 2007 banning incandescents) accelerated LED adoption but did not cap total lumen consumption. When municipalities save money per fixture, they install more fixtures and light areas previously left dark. Dark-sky ordinances exist in a few jurisdictions but remain rare and poorly enforced. All current policies target per-lamp efficiency (lumens per watt); no governance mechanism targets total lumen output or total lighting energy at the jurisdiction level. Fouquet & Pearson's UK data (1800–2000) show the real price of lighting fell 3,000-fold while consumption increased 40,000-fold, with income and price elasticities reaching 3.5 and −1.7 — confirming structural backfire across the entire modern lighting era.
What would unlock progress
Progress requires policy instruments that target total lighting energy or total lumen output rather than per-unit efficiency — analogous to emissions caps vs. per-vehicle standards. Approaches could include regional lumen budgets (total outdoor lighting limits), spectral requirements restricting blue-rich emissions in outdoor applications, or pricing mechanisms for light pollution externalities. Ecologically informed lighting standards (warm-white, shielded, dimmed after hours) exist technically but lack regulatory adoption.
Entry points for student teams
A team could instrument a campus or neighborhood to quantify the LED rebound effect: measure total lumen output and energy consumption before and after an LED retrofit, documenting whether and how lumen consumption increased. Alternatively, a team could design and model a lumen-budget regulatory framework for a municipality, drawing on carbon-cap analogs. Environmental engineering, policy analysis, and lighting design skills apply.
Genes
Structural cousins: stuck the same way, in other fields
Sources
Tsao, J.Y., Saunders, H.D., Creighton, J.R., Coltrin, M.E. & Simmons, J.A. (2010), "Solid-state lighting: an energy-economics perspective," Journal of Physics D: Applied Physics 43(35), 354001; Kyba, C.C.M. et al. (2023), "Citizen scientists report global rapid reductions in the visibility of stars from 2011 to 2022," Science 379(6629); IEA Lighting Tracking Report (2023), accessed 2026-02-23 go to source ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
This is a textbook Jevons paradox / 100% rebound case. The "problems of success" structure is explicit: the mechanism that succeeded (reducing cost per lumen) is the same mechanism that causes the harm (enabling massive lumen consumption growth). Related briefs: digital-datacenter-cooling-energy-intensity (different problem — cooling technology, not rebound). The Tsao et al. (2010) paper is the canonical reference for the lighting rebound; Fouquet & Pearson provide the 200-year UK dataset.
Reconciliation 2026-09-29: Fact repair after a site review read this write-up against its sources. Two sources were read in full on 2026-09-29. The Kyba et al. paper was read in the authors' accepted version, including its supplement, from the GFZ repository (https://gfzpublic.gfz.de/pubman/item/item_5014827_3/component/file_5014851/5014827.pdf); its citation was checked against the Crossref and Europe PMC records for doi 10.1126/science.abq7781. The IEA lighting page was read from an archived copy dated 2026-02-22 (https://web.archive.org/web/20260222131432/https://www.iea.org/energy-system/buildings/lighting), because the cited address now redirects to the IEA's general Buildings page. The first source could not be read in full: the publisher's page is closed and no open copy was found, so only its abstract and citation record were fetched (https://api.crossref.org/works/10.1088/0022-3727/43/35/354001, and the OSTI record at https://www.osti.gov/biblio/1380529). (1) The Source line gave the first source as Tsao and Saunders, 2012. The record for that title gives five authors (Tsao, Saunders, Creighton, Coltrin and Simmons), the year 2010, and Journal of Physics D: Applied Physics, volume 43, issue 35, article 354001. The Source line and the mention in the paragraph above now follow the record. (2) The text said the sky-brightness figure came from "51,351 participants across 19 countries." The paper counts 51,351 observations, not participants, and gives no count of countries. The sentence now says 51,351 observations. (3) The text said satellites show a smaller increase "because satellites miss the blue-light component of LEDs." The paper gives two reasons: the change in the colour of the light, and the fact that satellites see little of the light sent out toward the horizon. The sentence now says "in part because." (4) The text said blue-rich white LEDs "are driving" light pollution. The paper says the effect of the switch to LEDs on sky brightness is unclear, and that stars are becoming harder to see "despite (or perhaps due to)" LEDs. The sentence now says "may be driving." Confirmed against the sources and left as written: 9.6% per year, and doubling in about 8 years, which follows from that rate; the years 2011 to 2022; about 2% per year from satellites (the paper gives 2.2% for 2012 to 2016); LED efficacy above 100 lm/W (IEA); lighting energy use rising in 2022, particularly in large emerging economies (IEA); and the citation for Kyba et al. in the Source line. NOT checked, because they are not in the two sources that were read and the first source could not be opened: the 15 lm/W figure and the 7-fold gain; the 85–90% fall in cost per lumen-hour; 0.72% of GDP over three centuries and five lighting technologies; the statements that the rebound is 100% and removes the climate benefit of LEDs; the statement that a 4000K LED is about 2.5 times more disruptive than high-pressure sodium; EISA 2007; the statements about municipalities adding fixtures and about dark-sky ordinances; and the Fouquet and Pearson figures (3,000-fold, 40,000-fold, and the elasticities 3.5 and −1.7), for which the write-up names no paper. These are left as written. Genes untouched.