A good way to summarize this empirical evidence is to examine Figure 2 which displays
the average amount of ratio stacking per Congress. The intercept in this figure is clearly
positive, indicating something of a “baseline” of ratio stacking. The clear downward
slope reflects the expectation of the first hypothesis, that ratio stacking would be greatest
when the floor is closely divided. For any given point in the figure, we would expect an
upward shift if the committee were a prestige committee or small in size. Likewise, if a
committee were not a prestige committee or had a relatively large membership, we would
expect a downward shift in ratio stacking.
[Figure 2 about here]
Figure 2 also allows us to reconsider the conventional wisdom on House committee party
ratios. It is evident that committee party ratios will not always mirror the party ratio of
the floor. Instead, committee party ratios vary systematically with the size of the
majority party floor advantage. Closely divided floors tend to produce committee party
ratios that are consistently to the advantage of the majority party, and very large
majorities tend to give a slight benefit to the minority party.
Finally, let us note that only a partisan theory of congressional organization provides the
expectations that are supported by Figure 2. It does appear that the concept of majority
party reliability has merit since majority parties engage in the system strategy of utilizing
redundancy of components to enhance the reliable performance of committees.
Distributive theory would not predict the types of systematic advantages to the majority
party that are seen. Information theory also is unable to account for the data patterns
shown here. Instead of a positive intercept or a downward slope in Figure 2, information
Convention