This study in PLoS Medicine reports that, based on estimates from publicly available data, drug manufacturers probably spend more money on advertising than they do for research and development.
In the words of the study: "From this new estimate, it appears that pharmaceutical companies spend almost twice as much on promotion as they do on R&D. These numbers clearly show how promotion predominates over R&D in the pharmaceutical industry, contrary to the industry’s claim. While the amount spent on promotion is not in itself a confirmation of Kefauver’s depiction of the pharmaceutical industry, it confirms the public image of a marketing-driven industry and provides an important argument to petition in favor of transforming the workings of the industry in the direction of more research and less promotion."
This study makes it clear that it is using estimates to reach its conclusions. The manufacturers don’t report this data, so the authors had to take the limited available information and do what they could. Obviously, they could be wrong.