It's hard to predict what effect "sender pays" would have had on email if it were introduced from day one. For a start, it would mean that sending email would

It's hard to predict what effect "sender pays" would have had on email if it were introduced from day one. For a start, it would mean that sending email would only have been possible between organisations that are capable of sending/receiving financial (micro?) transactions between themselves. Such layer 8/9 considerations might have made the spam problem moot by making email itself moot. If you accept the premise of introducing some degree of financial friction into the email system, though, then I will mention a modest proposal of requiring new email-sending domains to put up good-behaviour bonds in a cryptocurrency that can be spent/burnt if the domain sends too much spam. (All existing domains would be grandfathered in, providing an incentive for them to agree to this system). The only controversial part would be deciding which entities had the power to judge whether a domain sent too much spam, but the obvious answer would be a consortium of the biggest email providers, since they already have the power to effectively render an email server unusable.