There are a ton of different factors to take into consideration here. Lets make sure we keep net GHGe and tailpipe GHGe separate here.
Net GHGe: Since 2015, satellite observation as well as international agriculture reporting data have started to show that the original assumptions surrounding Indirect Land Use Change (ILUC) for generation I and advanced biofuels have been grossly over calculating the effects of ILUC, as well as under and over accounting in other areas. Utilizing new data, there are several studies all pointing to a considerably lower GHGe footprint of Gen I and advanced biofuels. The most comprehensive study I've found to date has been A Life-Cycle Analysis of the Greenhouse Gas Emissions of Corn-Based Ethanol
The tl;dr here is, the current EPA model that claims ethanol lifecycle emissions are a 20% reduction from baseline gasoline still has the incorrect ILUC numbers. The real number as calculated in the above as well as several other studies [ GHGe recent evidence | ILUC data review ] is likely between a 35% and 50% reduction (with many studies finding in the 40%-45% range) from baseline gasoline. This is the net carbon accounting of EVERYTHING - Tailpipe, ILUC, vapor emissions, fuel production and transport, fertilizer - everything. It should be noted that the GHGe intensity of gen 1 and advanced biofuels are continuing to decline and will likely do so as shown below, and that none of these numbers even take into account other sources of ethanol with lower GHGe, including sugarcane and cellulosic.
From left to right: 2005 gasoline baseline (self explanatory), EPARIA: 2022 (current EPA model), ICF 2014 (correction for ILUC and other accounting from EPA model), ICF: 2022 BAU (business-as-usual projection to 2022 showing addition reductions), ICF: 2022 Building Blocks (described in paper, this is a theoretical potential based on new technologies such as combined heat & power, fiber digestion, etc). Note the dramatic difference in production & ILUC numbers
The first thing you'll notice in the above graph is that the tailpipe emissions of EtOH are a sliver of gasoline. This is confusing until you read the lifecycle analysis report.
The carbon dioxide emissions from corn ethanol are assumed to not increase the atmospheric CO 2
emissions as the biogenic carbon emitted is offset by the carbon uptake of new growth biomass. Life-
cycle CO 2 emissions from biofuel tailpipe combustion are not included in the analysis. The biofuel
tailpipe combustion CH 4 and N 2 O emissions are included however. These emission factors are based on
EPA’s MOVES model results (EPA, 2015b; EPA, 2010g).
This may seem "shady" if you aren't used to the carbon accounting, but every atom of carbon (no matter the form) involved in the actual transport biofuel comes from sequestration of an identical carbon atom in the air. This is an even, 1-1 exchange. They note that other compounds that are not CO2 are included, as they may be a 1-1 carbon exchange but they may have an uneven impact (GHGe).
Now, you were asking about the actual tailpipe emissions, not the tailpipe emissions that were matched with carbon sequestration via agriculture. For this we have to do a bit of math, and use the correct numbers. Reference this same study:
About 19.64 pounds (8.91 kg) of carbon dioxide (CO 2 ) are produced from burning a gallon of gasoline that does not contain ethanol.......About 12.73 pounds (5.77 kg) of CO 2 are produced when a gallon of pure ethanol is combusted
But as always when utilizing ethanol - gasoline comparisons, the per-gallon numbers don't mean anything. Using an energy density of 121.12 & 79.86 MJ/Gal for Gasoline & EtOH respectively, we end up with 73.56 & 72.25 gCO2e/MJ for Gasoline and EtOH, respectively. that difference is effectively nothing, as if you jumped between the LHV and HHV of gasoline you can flip which fuel is technically on top. We should expect this too, because we're talking about general hydrocarbon bonds being taken from a fixed state of reduction to completely oxidized, and regardless of the species we should see similar if not the same amount of energy released from breaking those bonds in the same way.
There are two caveats here.
The first: The above Net GHGe analysis takes into account conversion of N2 into N2O, and CH4, CO etc emissions at the tailpipe. The tailpipe GHGe analysis does not. Generally speaking, Ethanol fuel results in lower or identical non-CO2 emissions than gasoline, though under certain circumstances N2O & acetaldehyde emissions can be slightly higher than gasoline, and certain toxic, organic and other ozone forming compounds can be slightly lower than gasoline.
The second caveat is that EtOH can be considerably more efficient in producing usable work on a per-MJ basis than gasoline, due to its highly desirable qualities as an internal combustion engine fuel (specifically octane rating, octane sensitivity & heat of vaporization). EERC found that in both flex fuel & non-flex fuel cars, ethanol has a dramatic impact on fuel economy when looked at on a per-BTU or MJ basis. This is effectively an increase in efficiency in internal combustion engines, especially undersized engines under heavy load. I personally tested this on my own vehicle (2014 GMC Sierra, 5.3 flex fuel) and found almost identical results to what EERC found.
My personal results:
I had previously calculated my per-mile carbon intensity using the life cycle analysis here:
Here I have adjusted the per-mile carbon intensity utilizing the above calculated tailpipe emissions numbers instead of the lifecycle analysis numbers:
I believe we're just getting the point where vehicles are capable of utilizing higher blends of ethanol combined with gasoline, and I think there is going to be a dramatic reduction of carbon footprint of the U.S. fuel supply as higher amounts of ethanol are blended in, and as auto-manufactures begin demanding higher performing liquid transport fuels. Ford is already testing out all horsepower numbers for their Ecoboost engines on 94-95 octane fuel, and many manufactures are finding the value in producing downsized-turbocharged engines.