Again, with the prevalence of lopsided votes in Congress, these figures, while not
overwhelmingly large, are nonetheless significant. Senators facing large increases in black
registration levels in the South were voting inconsistently as a reaction to the massive change in
their constituencies. This massive change in voter registration levels among African-Americans
affected senators’ levels of uncertainty as to the preferences of their electors. The “erratic”
behavior characterized by inconsistent voting patterns among Democrats can thus be interpreted
as a rational response to changes in the electoral environment. The preferences of African-
American voters are better established today given their support for Democrats and their
relatively well-known preferences on public policies (Canon 1999; Swain 1995; Tate 1993).
However, during this time of flux in the U.S. South, the preferences of African-Americans were
not necessarily known to incumbent senators, especially those preferences that map onto the first
ideological dimension. While most newly enfranchised voters would likely lie somewhere in
either the left or the center of the first dimension, the southern Democratic party had been openly
hostile to African-American interests leading up to the time periods with large increases in black
registration. Further, preferences on non-civil rights issues among African-Americans in these
states were likely not well-known by senators during the 1960s: political polling did not focus on
African-Americans’ views at this time and senators also did not have an understanding of the
ideological preferences of a group that had been disenfranchised.
Note that in Model 2, neither the coefficient for black enfranchisement nor the interaction
is significantly different from zero. This suggests that the effect of the enfranchisement of blacks
on legislative inconsistency was dependent on which party dominated the state, rather than on
competitiveness alone. This result stands in contrast with that of the following analysis on the
larger sample. The results may also be explained by the time period examined. With this
28
Convention