CEO compensation has always been the subject of heated discussions. The main contention is with regards to the significant disparity between the salary of the average worker and that of a CEO; a typical CEO earns many times more than the typical worker. While the position of CEO is one of great significance, it is hard for one to accept the notion that a corporate executive would be astronomically more productive than the average worker.
As an economist, I tend not to make inferences based on raw numbers. Economists don't really care how much a CEO is paid as long as his or her performance justifies the huge salary. Let say one CEO's actions are found to have improved his company's profits by $1 billion. I have no problem with the company paying the person tens of millions of dollars a year in compensation; his pay is well-justified.
Unfortunately, measuring a CEO's performance is an extremely difficult task. Business is highly cyclical with performance peaks and valleys. More importantly, luck (both good and bad) plays a critical role in the success or failure of a business. Case in point, the swine flu epidemic brought booming sales of face masks and flu vaccines. As this is more or less a random event, one can't really attribute the increased sales and revenues to the performance of pharmaceutical CEOs.