7
affiliation of the non-Hispanic and non-Black Caucus (NHNB) Democrats is the population of
interest. Party affiliation is measured as a dichotomous variable with Republicans, Hispanics,
and Black caucus members coded as "0" and NHNB Democrats as "1"
4
. The expectation is that
NHNB Democrats are more likely to support Hispanic issues than Republicans.
Members of the House are expected to support Hispanic issues as the percent of the
district that is Hispanic increases. All of the Hispanic members of the House are elected from
districts with large Hispanic populations. To separate the impact of the percent Hispanic of the
district from the Hispanic representatives, the percent of the district that is Hispanic is for the
non-Hispanic representatives, and is coded "0" for the Hispanic representatives (Santos and
Huerta 2001).
A final measure to consider is ideology. Are the more conservative members of
Congress less likely to support Hispanic issues? The expectation is that the more conservative
the member of Congress, the less likely they are to support Hispanic issues (Santos and Huerta
2001). Ideology is measured using the American Conservative Union (ACU) Scorecard for the
107th Congress
5
. The ACU produces a scorecard for each year of a Congressional term and the
average ACU score of 2001 and 2002 is used as the measure of ideology. A higher ACU score
indicates a more conservative member of Congress. The highest possible score is 100 and the
lowest 0.
The correlation between the dependent variable and the ACU ideology scorecard is -.96,
thus it is inappropriate to use the ACU scorecard as an independent variable. Instead, two
dummy variables are created to analyze the role of ideology. One dummy variable codes all the
NHNB Democrats with an ACU score greater than or equal to 40 as "1" and all other cases as
"0". Another dummy variable codes all NHNB Democrats (plus the one independent) with ACU
Convention