Presenting: An Even Better Carbon-Tax (Spreadsheet) Model

Today we unveil our enhanced carbon tax model. The link on our Home Page takes you automatically to the new version. (You may also download the model by clicking here.)

As before, the model is a compact (600 kB) Excel spreadsheet that runs on PC’s or Macs with Excel 2003 or later. Like prior versions, it splits the U.S. energy-economy into a handful of “sectors” (e.g., electricity, passenger vehicles) and, for each, generates 20-year-or-more projections of usage and per-unit carbon emissions with and without a price on carbon pollution. You can set the prices yourself — we say “prices,” plural, because you can specify both the initial price and the rate of year-to-year increase; or you can use our “out of the box” pre-set prices.

The result is a pair of emissions projections — one with the carbon price and the other without — with both referenced to the same set of anticipated changes in economic activity and energy prices. The difference between the respective projections is the emission savings (reductions) imputed to the carbon tax.

Here are the key new features (these bullet points have been updated to reflect the 2014 version of the model):

• The model is baselined to 2013 levels of economic activity and fuel use. Thus it reflects the dizzying shifts of late in electricity generation shares, from high-carbon coal (which lost six percentage points of market share from 2010) to lower-carbon gas (up four points) and zero-carbon wind (up four points).
• The model employs “official” U.S. forecasts of economic growth, general inflation, and, most importantly, of prices for electricity, natural gas and crude oil, rather than our own forecasts. (The carbon tax is layered “on top of” the official prices.)
• Explicit treatment of inflation. First, the carbon tax itself can be indexed to general inflation, or not; second, energy prices (including the carbon-tax-affected prices) are translated into the nominal (inflation-inclusive) prices in which end-users actually experience (i.e., pay) them. While this may sound complex, the model results are more rigorous (accurate) than before.
• The model has seven sectors, up from five. We’ve split our catch-all category of “Other” energy uses (other than electricity, driving and other personal ground travel, freight movement, and aviation) in two: uses fueled by natural gas, and uses fueled by petroleum products. This enables the model to more fully capture differences in gas and petroleum prices including sensitivities to carbon taxes. A “miscellaneous” category captures CO2 emissions from non-electricity-generation uses of coal; from non-energy uses of fossil fuels such as natural gas used in chemical plants, LPG (liquid petroleum gas), lubricants and naphtha; and fuels used for energy in U.S. territories (which are not included in other tabs).
• We’ve overhauled our derivation of the tax’s impact on petroleum usage. Not only was our previous method needlessly convoluted; it also over-calibrated oil’s shares of the different sectors and thus led to overstating the reductions in petroleum usage from the carbon tax-caused reductions in future energy usage. (The predicted reductions are impressive nonetheless: for the Larson bill, which we model with inflation-indexed prices of $15/tonCO2 in the first year, incremented by $12.50/ton each year, the tenth-year reductions in oil usage are 4.3 million barrels a day from 2005 levels, and 2.6 million b/d from projected oil requirements without a carbon price; for comparison, the Keystone XL pipeline is intended to deliver 0.83 million barrels a day of crude from Canadian tar sands to U.S. refineries.)
• A new spreadsheet tab, Index, has links that improve navigation among the 22 different tabs.
• Clearer graphs of CO2 reductions and petroleum savings, and a new graph of revenue generation, expressed both nationally and per-household.
• Generally improved layout and presentation.

The key result — the model’s estimate of CO2 reductions from U.S. carbon taxes — is largely unchanged from earlier versions. In the tenth year of a “Larson”-type carbon tax, with the carbon tax rates noted in the large bullet-paragraph above, projected U.S. emissions CO2 from fossil fuel combustion are 1.8-1.9 billion metric tons less than predicted without a carbon price, a 33% reduction.

A few tabs in the spreadsheet aren’t yet complete. Nevertheless, the features enumerated here constitute such a large improvement over the prior version to warrant posting it today. We invite all users of the model, old and new, to kindly:

• Update earlier findings you may have drawn from prior versions of the model, substituting results from this one.
• Walk through the model with an eye toward evaluating it for logic, rigor, presentation and functionality. Please e-mail us all suggestions and criticisms, via info@carbontax.org.

Happy modeling, and best wishes.

   — Charles