RFF Comparison of EICDA and several other bills' provisions

Thanks David Kline‍ for recommending RFF's new resources. Other outfits like CGEP have also created helpful resources like this New Explainer: What You Need to Know about a Federal Carbon Tax in the United States.

The best place for CCL's resources in comparing these different price trajectories is the Carbon Pricing Bills In Congress page - including a PDF document outlining the major provisions for all major carbon pricing bills currently in Congress as well as a two-page PDF comparison table and charts. 😀 

PS for anyone interested in subscribing to Resources For the Future's excellent newsletter for future updates and studies, go here: https://www.rff.org/subscribe/

Thanks, Brett Cease‍ that's a great resource, and a "living document" that gets updated as new bills come in. As of this writing, it already has the Whitehouse-Schatz bill just introduced last week.

I am making a graph compaing all the price paths of the bills on that list. First question: has that wheel already been invented and if so where can I find it?

Second more technical question to make sure I get the calculations right. A couple of the bills add a constant amount to the price, adjusted for inflation. The summary of the Deutch bill says, for example "Rate of increase: $10 annually above inflation."

Here's my interpretation of what that means, which I'd like to check: Assume 2% inflation for purpose of illustration. First year Deutch price is $15. At the end of the year, you add $10 inflated, or $10 (1+.02) = $10.20. The next year you inflate again, adding $10*1.02*1.02 = $!0.404. (I suppose to be really accurate, one might round to the nearest penny before inflating again. But never mind that.)

So do I have that right?

Thanks!

David

Hi David,
Very sharp of you to spot that question. Actually, in Research it came to our attention that the 2019 bill, H.R.768, was written in a way consistent with what you are saying: that only the annual increase is adjusted for inflation. That would make seriously weaken the price impact over time because, for example, after 5 years (assuming 2% inflation) you would have $15 + $10.20 + $10.20 + $10.20 + $10.20 = $55.80, which in inflation-adjusted dollars would only be equivalent to $55.80/(1.02^4) = $51.57. We inquired with the Deutch office whether that was really their intention and confirmed that it was not. So a small text change was made for 2021, and now the inflation adjustment applies to the cumulative amount so that in Year 5 the total in inflation-adjusted dollars would be $58.27 (calculation is hard to show).. 
You can see that this impact grows substantially as the years go by. In Year 10, the total in Year 10 dollars would be $117.42, versus only $106.80 with the old language.
This change is not earth-shaking, but at least it makes the price a little stronger, especially if inflation ticks up in some years, or the annual increase ticks up to $15 for missed targets. 
Rick

Hi Rick,

I'm having trouble following what you're saying about how it works in the current bill. I don't know  how to interpret "the inflation adjustment applies to the cumulative amount." I couldn't come up with an interpretation of that in the spreadsheet that got the answer 58.27 in real dollars in year 5.

One interpretation would be to take the nominal series 15, 25, 25 ... and inflate each one of those. That's obviously not what you mean, because then the adjusted price in year 5 would be exactly 55. I'm not seeing how you get something larger than that.

You could send me a small spreadsheet that calculates the inflation-adjusted prices in the EICDA and I'm sure I could decode it. The Durbin bill is the other one that uses an annual dollar increase. Does it work the same was as EICDA except with a $25 first-year price?

All clues gratefully appreciated. I will happily share the price comparison chart that results.

Best,
David

 

David,
Sorry about my not-so-great explanation. I went back and re-read the bills (again) and had to make further minor adjustments to account for the reference years for inflation in each bill text. That changed my numbers slightly, but not significantly. In Year 5, the inflation-adjusted price under H.R.763 would be $56.00, while under H.R.2307 it would be $58.82. In terms of purchasing power (compared to Year 1), they would be, respectively, $51.74 and $54.34.
I'm attaching a spreadsheet to show how this works. The complications are that there is no inflation adjustment in Year 1, but in Year 2, the price is adjusted for TWO years of inflation because of the way the text is written. Then in subsequent years, inflation is applied differently depending on the bill language..
The spreadsheet also shows the Durbin-Newman bill, which follows the same method as H.R.2307, except for the higher Year 1 starting price of $25.

Rick

I read the relevant sections of both bills. I think I'm with you now. EXCEPT: I believe that the $58.27 figure you quote for the 2021 bill in year five is actually the NOMINAL dollar figure (i.e. dollars current in year 5.) I get $53.84 = 58.27/1.02^4 for the inflation adjusted price (i.e. dollars current in the starting year). It seems to me impossible that the inflation-adjusted figure could be higher than $55, the year 5 price if nothing were inflated.

I'm imagining that the authors really wanted the inflation-adjusted price series to be 15, 25, 35 ... This doesn't quite get us there. The difference is relatively small, but would get big if the program went on for a very long time. But on the couple-three decades time frame it doesn't make a big difference.

I think we're on the same frequency here -- I believe 'nominal' dollars is right, I may be garbling the terminology. See the spreadsheet, where I use "inflation adjusted" dollars to mean the same thing, which is the dollar amount that would be shown on a receipt in any given year. 
Rick

Thanks a million for the spreadsheet. I agree that we're on the same page. I'm not quite sure where inflating the first increment by 2 years comes from in the bill, but I'll take your word for it on that on that and do it that way in my own spreadsheet.

Great,
The 2-year thing comes from the basis year that each bill identifies for CPI calculations. In H.R.763, it was 2018. In H.R.2307, it's 2020 (or should be - the text says 2010, but I'm sure that's a drafting error that will be corrected). The Durbin bill says 2022, but the carbon tax doesn't kick in until 2024. So in every case, it's 2 years.
Rick

With great help from Rick, I made a chart comparing the prices in the EICDA, Durbin, Whitehouse-Schatz, and Fitzpatrick bills. Here's what it looks like. The prices depend on the assumed inflation rate, and I assumed 2% for the CPI (and if you're paying closer attention than may be worthwhile, 1.75% for the chained CPI used in some of the bills.)

I did not include the Climate Leadership Council proposal both because it's not an actual bill, and because it's not specified as precisely as a bill needs to be--not to mention that they keep changing it. Their current version falls between the Whitehouse and Fitzpatrick prices.

There are two price trajectories for Fitzpatrick because it eliminates the federal fuel excise tax, lowering fuel prices. The lower curve gives an approximation of the net effect given that tax decrease, on average, in the transportation sector.

I'm happy to share more details, e.g. the spreadsheet, if anyone is interested.

Hi @David Kline 

Can you specify what year EICDA price first exceeds Whitehouse's SOFA price?

Also, I'd love a copy of the spreadsheet if you want to post it here.

Thanks,
Tony

 

@Tony Sirna – 

2027 is the first year that EICDA prices exceed the Whitehouse-Schatz prices. Note on that: one way to compare price series to each other is to take their present value. That measure answers the question "If I had a widget that saved me 1 ton's worth of carbon fee each year, how much would it be worth to me over the 20-year time horizon shown in the spreadsheet? By that measure, EICDA prices are the highest, Durbin is very slightly lower, and Whitehouse is about 15% lower.

I'm happy to share the spreadsheet, so find it here:

https://drive.google.com/file/d/12o5Rh1zm5_eEZGYv8sgAwbFg5w3cEOjs/view?usp=sharing 

It is open for comments but not for editing. You can download it if you want to change something. Note for example that the prices all depend on the assumed inflation rate--in fact two different assumed inflation rates. They change slightly over reasonable ranges of inflation assumptions, but stay in the same general relationship to each other.

As I mentioned, I worked with Rick Knight on this and he was an enormous help. Any errors are my responsibility. And there may very well be some--it's easy for errors to creep into Excel, and there are a zillion little picayune differences among all the bills. No two of them calculate their prices exactly the same way, although EICDA and Durbin are close. Finally, I am not 100% sure that Rick and I are interpreting all the bill language correctly. I believe he's working on getting verification on that.

All of that said, this is the working version for now. If we make any changes, I'll update the spreadsheet in that location and make a note of it here.

@Rick Knight,

Comparing the prices in my post above to the  figure below from your recent blog post that estimates the emissions reductions from various bills raises a couple of questions. For one, although the EICDA prices (at least as calculated above) are higher than Durbin, Durbin reduces emissions more. Also, Whitehouse prices are lower but emissions reductions are essentially the same as EICDA. Do those differences come from other aspects of the bills besides prices that you modeled in your analysis? If not, let's compare our price calculations.

Thanks!

 

 

 

This chart is from the Energy Policy Simulator. I'm not sure why the emissions cuts are a bit more from the Durbin and Whitehouse bills, but the differences are quite small. It may be due to the higher starting prices, or due to the fact that the simulator doesn't accommodate a steady percent escalation so I had to express it in fixed amounts at selected years. Maybe adjusting these amounts for every single year from start to finish would give a slightly different result. Otherwise, the only differences are in Government Revenue Accounting. This part of the model is very hard to interpret, but it should not affect emissions. 

OK, we can chalk up those apparent contradictions to the approximations you made to the price paths. I would like to write an article pointing people to your blog post comparing the emissions, also including my price chart. Let me know if there's a problem with me using the figure from your post that I included above.

Updated price comparison including the Larson bill. A careful reading of the bill language reveals that the Larson prices are not directly comparable to all the other bills. That is because the Larson bill assesses its fee at the point of first sale, rather than at the point of extraction or import. I chose to ignore that, assuming that the difference in the effective fee would be small.