I think Dr. Mankiw's advice is cogent. In one of his recent posts, he mentioned that a student wondered whether he should pursue a PhD in economics:
"Would you recommend someone to do a PhD if he knows he's unlikely to become a star in the field (weak math background due to lack of trainings and not being particularly gifted; lack of confidence in his creativity and talents) and does not have a burning desire to do research, but has interests in social science, enjoys learning, and likes to be able to interact with people he admires and respects? Or do you think it's better for him to work first until he's certain that research is what he wants to do?"
Here is Dr. Mankiw's advice:
A PhD takes quite a bit more time and concerted effort than most graduate degrees. An MBA is two years, a JD is three years, while a PhD is often about five or six years. This fact has a couple implications. First, you should be more confident that you really want the degree before you start (although there is nothing dishonorable about starting a PhD and then changing your mind after a year or two once you recognize that it is the wrong path for you). Second, you should not take off much time after college before starting. A year or two is fine, but more than that can be problematic, for the simple reason that as most people approach age 30, their willingness to lead a student lifestyle diminishes.
If one has the requisite degree of enthusiasm and commitment, however, one needn't be a superstar to pursue a PhD. A person can be perfectly happy with a PhD from a lower ranked school, followed by a career as a college teacher. There are thousands of economics professors around the country (as well as PhD economists in government and the private sector), and most lead very satisfying lives without ever being candidates for the Nobel prize. The one thing they share is a passion for the study of economics.