Some of the reasons I heard are:
* it is difficult to understand that donating for technology innovation is really a donation, it feels more like investing in companies. 
* policy advocacy sounds like lobbying
* it is kind of abstract if one compares it to other donations more widely known like planting trees; or donating medicines, vaccines or books to improve the health of the poorest.
* the impact of the donation is uncertain and based on estimations only.
* such a donation does not give this inner glow/good feeling that they expect to get when making a donation

Thanks for the input, henrith.

This anecdotal evidence from Sweden that you mentioned is what I also noticed when talking to people interested in climate change, but not into the EA-movement.

Regarding Giving Green, there was a very interesting discussion in the forum.

It seems like the differentials between BURN and CATF are more than 10x and could be even 100x, if CATF eventually managed to have an impact of $0.20/ton. This does not seem unrealistic considering that the estimations of $1/ton are conservative.

The study I referenced about Eden Reforestation mentioned an estimated impact of $0.36/ton. This would actually be in the same order of magnitude as the conservative estimates for the most effective organisation. This is why I mentioned that organisation as a potential alternative.

An additional point to consider is that it might not be the same:
 * making a personal recommendation to someone that might probably not donate otherwise.
* recommending it on a website. 

If an organisation is recommended on a website, there is the risk that people that would otherwise donate to the most effective organisations will change their donation to those less effective ones, having a relative negative impact. If I remember well, this was one of the arguments Johannes used, which I found fair enough, especially if we are talking about orders of magnitude of difference.

Looking forward to hearing more thoughts on this topic :)

I am also curious about how would an organization like Eden Reforestation Projects rank when also considering that they are giving jobs to extremely impoverished people, so even if it is not a direct transfer of money like GiveDirectly does, it might be comparable and some of the benefits could also apply.

Great report, thanks!

Note that the links to impactmatters.org are broken since they now belong to Charity Navigator. 

Here the articles you mentioned but retrieved from archive.org:

+ Overview

+ Methodology

Also, I have a question regarding the following point:

"Once a forest is mature, it does not continue to absorb carbon. Instead, trees decay and grow in roughly equal measure, with the mass of the forest (and therefore the carbon sequestered) remaining roughly constant."

I've seen that statement often, but carbon is stored in the soil as well, isn't it? However, I have no clue about which % of the carbon absorbed by a tree remains in the soil and which % is released again when it dies. 

How did you approach this?

What is your view on CoolEarth? It is not an advocacy charity but the cost per ton was in past reports similar to the advocacy ones (even if those are conservative estimates).

I liked the approach ' "how much more money does this crowd in?" compared to "how much does this crowd out from high impact options? " ', but in this case, the difference is not as big as with offsetting, so I am not sure what would be the outcome.

Also, is there any report or article where you explain in more details the revision of your view on CfRN?

I understand what you mean. Carbon offsetting is something many people is already familiar with and also tangible (direct action).  Indirect actions like developing new technologies, including advanced nuclear, or lobbying to influence political solutions,  might be the most effective way, but it is less compelling to many people. 

I used to offset my carbon footprint as well until recently. It took some time for me to absorb the facts and act consequently. However, I still find the top charities currently recommended by Founders Pledge kind of abstract and "unsexy".
These adjectives might not be considered relevant for advanced effective altruists, but I think the emotional component should not be neglected. This is especially important when talking to other people about it. 
It is only my own experience and intuition, but the current top charities do not make one feel excited about it and eventually many will not donate anything, while if other charities that perform direct action are mentioned, it could be more compelling. 

The compromise I found is CoolEarth. Protecting forests is tangible (direct action) and people generally like the concept of protecting forests and can grasp it instantly. In addition, according to the studies available (see below - unfortunately, I do know of any more updated versions) we are talking about the same order of magnitude when comparing it to the top charities.
Therefore, when talking to other people, I usually start talking about CoolEarth and only in cases where I see they are interested in getting deeper I will talk about the other options.

Regarding the 100x effectiveness:
* Clean Air Task Force: $0.10-$1/tonne of CO2e. Source: Founders Pledge report 2018
* Coalition for Rainforest Nations: $0.02 - $0.72/tonne of CO2e. Source: Founders Pledge report 2018
* CoolEarth: $0.18-$0.71/tonne of CO2e. Source: GWWC report 2016 (Note: for this figure, both directly and indirectly protected areas are counted)

I find this topic very interesting and would love to hear other opinions and arguments :)

Very nice project!  I'm really looking forward to seeing which athletes take the pledge.
Congrats and best of luck! 

I have a small question: I was surprised to see Atmosfair listed as a recommended charity.
They definitely do a good job in the offsetting field, but they are 100x less effective than other charities included like Clean Air Task Force, Coalition for Rainforest Nations or even other unlisted charities like CoolEarth.

I'm curious to learn more about why you decided to include them anyway?

What about creating a list in GitHub? Does it make sense?
Anyone can suggest modifications there + it could be directly used as a source to create interactive tables...