I found this paper while looking up some other information in another post (referenced in the post, but not posted explicitly):
Emission reductions by the USA in 2020 and the risk of exceeding 2°C warming
This report examines the relationship between the level of emission reductions to be undertaken by the United States by 2020 and the risk of exceeding a 2° Celsius (3.6° Fahrenheit) warming globally in the coming century, within the context of a new international agreement on climate change to be adopted at Copenhagen in 2009. International negotiations are focusing on a reduction range for Annex I countries as a group of 25-40% below 1990 levels by 2020 and for Non Annex I countries of a 15-30% reduction relative to business as usual growth by the same time.
The level of action by the United States is a very significant political variable in the Copenhagen climate negotiations and is likely to influence the level of ambition for the entire agreement. A stronger level of action by the USA would likely lead to more action from others, and vice versa. A delay in achieving emission reductions consistent with the 25-40% Annex I reductions would likely lead to delay by others. It is in this context that the Administration of President Obama faces difficult dilemmas. Under the previous Bush Administration very little was done to limit greenhouse gas (GHG) emissions, and by 2006 they were some 14% above 1990 levels, making a reduction to the 25-40% below 1990 levels within little more than a decade a very difficult task. President Obama has indicated that the United States should reduce its domestic emissions to 1990 levels by 2020, and to 80% below 1990 levels by 2050. Recently the US Special Climate Envoy Todd Stern argued that meeting the 25-40% reduction target range from 1990 levels by 2020 for the Annex I countries can be deferred, and faster reductions in the post 2020 period can make up for slower reductions to 2020.
From the conclusion:
Delaying emission reductions by the Annex I group by 10 years, from 2020 to 2030, results in significantly higher cumulative greenhouse gas emissions and increases the rate of emission reduction in future decades. The probability of exceeding 2°C warming is increased by about 15% for such a delay, from a base probability for the two non-delay scenarios of 14% (6% to 32%) and 27% (14% to 48%), respectively. Given that it is likely that should the US take weaker action, others would follow suit, a delay in achieving significant reductions by United States would thus likely result in an increased probability of exceeding a 2°C warming.
As a consequence, the assertion that a delay in reducing emissions can be made up by steeper reductions in later years is not supported by the analysis presented here.
The increased risk of exceeding 2°C caused by a ten year delay in reaching a 30% reduction levels for the Annex I countries is comparable to the increased exceeding risk caused by a 35% higher global level of emission in 2020. This analysis shows that a delay in reducing greenhouse gas emissions by ten years causes a higher overall level of emissions in 2020 and converts an emission pathway with about a 1 in 7 chance of exceeding 2°C to a 1 in 4 chance. Arguably this is a significant change in risk for a task that is already very difficult (Schellnhuber 2008) and it certainly does not support the argument that steeper reductions later can make up for a delay in reducing emissions.
Link to paper (PDF):
http://sites.google.com/a/climateanalytics.org/test/variousfiles/USAemissionreductionsandriskofexceeding2oC-March2009_EditedVersion2-3.pdf?attredirects=0