Saved analyses
Annual Washington-specific gasoline price change vs. 2015-2022 baseline
Question: How much more did Washington regular gasoline cost in 2023 through 2026, relative to its 2015-2022 relationship with each comparison region?
Result: 2025 effect: 41-56 cents per gallon vs. comparison regions without fuel carbon pricing; 2024: 20-38 cents; California comparison 3-19 cents.
Operation: difference-in-differences-annual
Scope: Five comparison regions by four periods (2023, 2024, 2025, 2026 through 2026-10-05).
View this analysis
- Difference-in-differences estimates capture all Washington-specific factors, not only the Climate Commitment Act; the Clean Fuel Standard (under 1 cent per gallon by Ecology's estimate), refinery outages, and other state-specific shocks are included.
- Placebo tests show single-year pre-policy deviations of up to about 20 cents per gallon, so any one year's estimate is approximate; HAC standard errors understate that uncertainty.
- California has its own cap-and-trade and Low Carbon Fuel Standard and is a weak comparison; it is reported, not relied on.
Quarterly measured premium and auction-implied carbon cost
Question: Did the measured Washington gasoline premium move with the carbon allowance price quarter by quarter?
Result: Quarterly measured premium and carbon cost per gallon, 2023Q1-2026Q3.
Operation: difference-in-differences-quarterly-and-carbon-cost
Scope: Fifteen quarters, 2023Q1 through 2026Q3.
View this analysis
- Difference-in-differences estimates capture all Washington-specific factors, not only the Climate Commitment Act; the Clean Fuel Standard (under 1 cent per gallon by Ecology's estimate), refinery outages, and other state-specific shocks are included.
- Placebo tests show single-year pre-policy deviations of up to about 20 cents per gallon, so any one year's estimate is approximate; HAC standard errors understate that uncertainty.
- Diesel carbon cost assumes no biodiesel or renewable diesel blending and is an upper bound.
Carbon cost per gallon vs. measured premium (chart series)
Question: Did the measured Washington gasoline premium move with the carbon allowance price quarter by quarter?
Result: 30 rows: two series by 15 quarters, cents per gallon in January 2024 dollars.
Operation: reshape-quarterly-comparison-long
Scope: Fifteen quarters, two series.
View this analysis
- Difference-in-differences estimates capture all Washington-specific factors, not only the Climate Commitment Act; the Clean Fuel Standard (under 1 cent per gallon by Ecology's estimate), refinery outages, and other state-specific shocks are included.
- Placebo tests show single-year pre-policy deviations of up to about 20 cents per gallon, so any one year's estimate is approximate; HAC standard errors understate that uncertainty.
Pass-through summary statistics
Question: How closely does the measured premium track the auction-implied carbon cost?
Result: Against the 7-state group and the U.S. average, the slope is about 1.0 and the mean premium (38.7 and 35.0 cents) is close to the mean carbon cost (38.9 cents).
Operation: correlation-and-ols-slope
Scope: Fifteen quarters, three comparison regions.
View this analysis
- Difference-in-differences estimates capture all Washington-specific factors, not only the Climate Commitment Act; the Clean Fuel Standard (under 1 cent per gallon by Ecology's estimate), refinery outages, and other state-specific shocks are included.
- Placebo tests show single-year pre-policy deviations of up to about 20 cents per gallon, so any one year's estimate is approximate; HAC standard errors understate that uncertainty.
- Fifteen quarterly observations is a small sample; correlation is descriptive and does not by itself establish causation.
Comparison with Ecology consultant (Greene Economics) estimates
Question: Does this analysis reproduce the state-commissioned 2023-2024 estimates?
Result: This analysis is within 6 cents of the Ecology consultant for EIA-based comparisons (2024: within 4); the consultant's AAA U.S.-average 2023 estimate (51.4) is higher than this EIA-based estimate (39.7).
Operation: compare-published-estimates
Scope: 2023 and 2024.
View this analysis
- Comparison groups and price sources differ (AAA vs. EIA; all states vs. seven states), so close but not identical values are expected.
Placebo test: fictitious 2019 policy start
Question: How large are Washington-specific deviations in years without the policy?
Result: Pre-policy single-year deviations range from about -21 to +21 cents.
Operation: difference-in-differences-placebo
Scope: Three comparison regions by four pre-policy years.
View this analysis
- Indicates the noise level of single-year estimates.
Sensitivity: 2019-2022 baseline
Question: Do results change with a shorter, more recent baseline?
Result: Results change by under 6 cents with the shorter baseline.
Operation: difference-in-differences-annual
Scope: Three comparison regions by four periods.
View this analysis
- Difference-in-differences estimates capture all Washington-specific factors, not only the Climate Commitment Act; the Clean Fuel Standard (under 1 cent per gallon by Ecology's estimate), refinery outages, and other state-specific shocks are included.
- Placebo tests show single-year pre-policy deviations of up to about 20 cents per gallon, so any one year's estimate is approximate; HAC standard errors understate that uncertainty.
Estimated cost to Washington on-road fuel users
Question: About how much did the carbon cost add to what Washington drivers paid for on-road fuel?
Result: About $3.8 billion for gasoline by both methods, about $5.1 billion including on-road diesel (allowance method), 2023 through 2026Q3.
Operation: gallons-times-cents
Scope: 2023Q1 through 2026Q3.
View this analysis
- Diesel figures assume no renewable blending and are an upper bound; diesel has no measured-premium estimate because EIA publishes no weekly Washington diesel price series.
- FY2026-FY2027 gallons are state forecasts, not actuals.
- Estimates are costs to fuel users, not state revenue; the two are not directly comparable.
weekly-prices.csv
text/csv · 54.4 KB (55,725 bytes)
