Through my current and past work as associate editor of several refereed journals, I have discovered a negative correlation between the number of papers that a scientist publishes per year and the number of times that that scientist is willing to accept manuscripts for review. In other words, the biggest consumers of peer review seem to contribute the least to the process.
There are two solutions to this situation. We could abolish peer review altogether, which would be tantamount to doing away with science as we know it. Alternatively, we can apply a well-known sociological principle, according to which no voluntary association can survive without incentives to increase compliance with its rules and penalties for disobedience. I therefore suggest that journals should ask senior authors to provide evidence of their contribution to peer review as a condition for considering their manuscripts. Such evidence should be easily verifiable in this age of data mining.
So, if you publish 10–20 research papers a year with the help of 30–60 referees, do your bit in return.