The issues of cyclical majorities and instability of collective choice
s have been in the forefront in the discussion of social choice mechan
isms. Cycling, lack of equilibria, and so called chaos theorems have b
een prevalent in the public choice literature. Whether cycling actuall
y occurs in processes that are decided by majority rule has been a lon
g-standing question. However, cycling has not been made quantifiable n
or has it been clear how one would determine empirically whether major
ities are cyclical. In this paper, cycling has been given an empirical
meaning. This paper provides a method to test for cyclical majorities
and applies it to decisions made in the US Congress. The test results
indicate stability and the presence of persistent winners and losers
in Congress.