I see two significant long-term trends, both of which are probably unstoppable:
1. In percentage terms, fewer and fewer economists are Americans by birth and upbringing. Non-Americans are less likely to be fully fluent in English, which encourages mathematics. Non-Americans also tend to be less market-oriented in their thought. In any case they are less likely to stand along traditional U.S. ideological fault lines or even share ideological fault lines with each other.
2. Empirical work has a shorter half-life than does theory, at least on average. Yet most people today, including at top schools, do empirical work. There will be fewer giants and more middling-size figures of temporary import. If you tenure a top empirical economist, what does your department look like when he or she is 60? What percentage of these people are skilled enough, or care enough, to continue reinventing themselves? Maybe some top departments won't look so top in twenty years' time.
I believe also that the higher relative status of empirical work favors the status of women in the economics profession.
Field experiments have a longer half-life than regressions, precisely because they are so costly to redo.