Individual events have effects (Erikson, MacKuen, and Stimson 2002; Kernell 1978).
Finally, some have theorized that all effects on approval are dependent upon media (Mutz
1993) or elite opinions (Brody 1991; Haight 1979). Again, it is quite clear that scholars
have expended a lot of effort in determining what affects presidential approval.
However, if one accepts the notion that most of the economy, media, and foreign affairs
are outside the control of the president or Congress, very little work has been done on
how presidential approval responds to events that can actually be affected by what is
done in Washington, D.C. An exception is Groseclose and McCarty’s (2001) work on the
negative effects of vetoes on presidential popularity, which studies a phenomenon for
aspect of the politics can impact political opinion. However, we still have very little
research on what the political consequences of legislating are. To use the feedback loop
analogy, we have yet to study the first link in the process: how does the public react to
major legislation?
What the studies above have in common is reliance upon aggregate data. The
relevant unit of analysis for most scholars of presidential approval has been the public at
large. What scholarship does exist on individual level approval is sparse and generally
constrained to one president. Some of Edwards’ work (Edwards, Mitchell, and Welch
1995) on issues and issue salience has made use of individual level data. Scholars have
looked at gender, age, education (Gilens 1988), and partisan (Ostrom and Simon 1988;
Tedin 1986) components of presidential approval. Only one work (West 1988) studied
people under different presidents. What we are left with is a lack of a basic
1
It is dangerous to assert that the president is solely responsible for a veto, as vetoes usually do not come as
a surprise to Congress, which passed the legislation in the first place.
4
Convention