Energy Demand & Emissions Outlook
(UN medium variant)
BAU 2023→2050
reduction potential (IEA)
vs. 1.5 °C path (2030)
GJ per $1 000 GDP
GJ per $1 000 GDP
World Population Trajectory (UN Variants)
Population Growth by Region 2023→2050
Urbanisation Rate
By 2050 ~68 % of the world will live in cities — urban households consume 40–70 % more energy per capita than rural ones but are more amenable to public-transit and district-heating solutions.
Age Structure & Energy Lifecycle
Typical per-capita energy by life stage
Peak consumption occurs 40–60: larger homes, more driving, higher discretionary spending. Ageing populations in advanced economies moderate aggregate demand growth.
Middle-Class Expansion
The global middle class is projected to reach ~5.4 B by 2030 (Brookings/Pew). Each new middle-class household adds ~3–7 MWh/yr in energy demand — primarily in South/Southeast Asia.
Demand Concentration: Where Does New Growth Come From?
Net growth in primary energy 2023–2050 by region (EJ)
Africa and South/Southeast Asia account for >80 % of global population growth to 2050 and an even larger share of net energy demand growth under BAU. The critical question is whether this demand is met by fossil fuels or leapfrogged to renewables.
Global Final Energy Demand by Sector (EJ)
Per-Capita Energy vs. GDP — Kaya Identity Components
The Kaya Identity decomposes CO₂ into population × (GDP/pop) × (energy/GDP) × (CO₂/energy). Decoupling requires the last two terms to fall faster than the first two rise.
Demand Drivers by Sector: What Matters Most
Residential (28 % of final energy)
Heating/cooling dominates (45 %), water heating (18 %), appliances and electronics (22 %), lighting (8 %). Behavioral interventions — thermostat nudges, appliance standards, social comparison — deliver 8–15 % reductions with zero capital cost.
Industry (37 % of final energy)
Steel, cement, chemicals, and aluminium account for two-thirds of industrial energy. Process heat >400 °C is the hardest to decarbonise. Circular economy and material efficiency could cut industrial demand 25–40 % by 2050 (IEA).
Transport (26 % of final energy)
Light-duty vehicles are electrifying rapidly. Aviation and shipping — 5 % of energy but 11 % of transport emissions on a 20-yr GWP basis — remain structurally difficult. Modal shift and urban form changes offer 15–20 % long-run savings.
Energy Intensity Trend (EJ per $T GDP)
Electrification Share of Final Energy
Electricity's share of final energy must rise from ~21 % today to ~50–55 % by 2050 in net-zero scenarios. Electrification changes how much primary energy is needed (efficiency gains) but not necessarily how much people consume in useful-energy terms.
Key Behavioral Mechanisms in Energy Markets
Present Bias (Hyperbolic Discounting)
Consumers drastically discount future energy savings. A $200/yr bill reduction starting next year is valued as if it were worth ~$60 today (implicit discount rate ~60 %). This explains the "energy efficiency gap" — underinvestment in cost-effective retrofits even at low interest rates.
Loss Aversion (Prospect Theory)
Kahneman & Tversky: losses loom ~2× larger than equivalent gains. Framing energy waste as a loss ("you're wasting $340/yr") increases conservation 18–30 % over equivalent gain framing ("you could save $340/yr").
Social Norm Effects
Opower/Oracle Utilities showed that comparing household consumption to neighbourhood averages reduces demand 1.5–3.5 % persistently — with no financial incentive. Effect is strongest for the heaviest users (above-norm households cut more than below-norm households increase).
Demand Reduction Potential by Behavioral Lever
Price Elasticity of Energy Demand
Short-run elasticities are small (−0.05 to −0.2) — people don't quickly change behaviour in response to price. Long-run elasticities (capital stock turnover) are 3–5× larger. Carbon pricing alone is insufficient without complementary behavioural nudges.
Rebound Effects
The rebound effect partially offsets efficiency gains: cheaper energy services lead to greater consumption. Household direct rebound averages 10–30 %; macro-economy-wide rebound can reach 50–100 % (Jevons paradox in some sectors).
Technology Adoption Curves: S-Curve Dynamics
Behavioral drivers of adoption acceleration
Social contagion: EV and solar adoption spreads spatially within social networks — a neighbour's EV purchase increases your probability of buying one by 0.5–2.5 pp (Graziano & Gillingham 2015).
Identity effects: Early adopters signal prosocial identity; once penetration crosses ~15 %, mainstream adoption becomes status-neutral and follows economic logic alone.
Infrastructure feedback: Charging networks and installer capacity are self-reinforcing — higher adoption raises infrastructure investment, lowering adoption costs further.
Global CO₂ Emissions Trajectory — Scenario Fan
Cumulative Emissions Budget Remaining
At current emission rates (~37 GtCO₂/yr), the 1.5 °C budget (~250 GtCO₂ as of Jan 2024) is exhausted in ~6.7 years. The 2 °C budget (~1 200 Gt) lasts ~32 years.
Population-Weighted Per-Capita Emissions: Where Do We Need to Go?
The convergence challenge
A fair 1.5 °C budget implies ~1.0 t CO₂e/person/yr globally by 2050. Current ranges are vast: US ~14.5 t, EU ~6.8 t, China ~8.1 t, India ~2.4 t, Sub-Saharan Africa ~0.7 t.
High-income countries must reduce 85–95 % from current levels. Developing nations must grow living standards without following the fossil-fuel industrialisation path.
Policy Instrument Effectiveness (GtCO₂e reduction by 2030)
Carbon Price Required by Scenario
The IEA Net Zero scenario requires $130/tCO₂ by 2030 and $250/t by 2050 in advanced economies. Current global average effective carbon price is ~$5/t — a 26× gap. Behavioral nudges reduce the carbon price needed to achieve a given outcome by 15–35 %.
Policy Portfolio: Combining Instruments for Maximum Impact
Pricing & Markets
Carbon taxes, emissions trading, feebates, congestion charges. Addresses production cost signals but insufficient alone given behavioral barriers. Most cost-effective when recycled as dividends (sustains political support, addresses distributional concerns).
Standards & Mandates
Appliance standards, building codes, EV mandates, fuel economy rules. Overcome present-bias and split-incentive problems. Deliver predictable demand reductions independent of consumer behavioral change — but can suppress innovation incentives if poorly designed.
Nudges & Information
Smart meters, social comparisons, default green tariffs, eco-labels, disclosure requirements. Low-cost, high-legitimacy, but modest effect sizes (1–8 % demand reduction per instrument). Most powerful when layered with pricing and as complements to infrastructure investment.
Demand-Side Mitigation Potential by Region (GtCO₂e/yr by 2050)
Cost Curve of Demand-Side Interventions
Many demand-side measures have negative abatement costs (pay back within the analysis period). Building efficiency retrofits, appliance standards, and modal shift are consistently in the top quartile of cost-effective climate interventions.