I am a Professor of History and Chair of the History Department at American
University in Washington, DC. My areas of expertise include political history,
voting analysis, and historical and quantitative methodology. My scholarship
includes articles and books on political history, quantitative methodology,
and the application of social science methods to voting rights issues. I
have worked as a consultant or expert witness for both plaintiffs and defendants
in more than sixty federal voting rights cases.
The corrected and uncorrected data for each state and for white and minority
groups is reported on the website of the United States Bureau of the Census:
>www.census.gov<. This website also provides corrected and uncorrected
data for units within the state, partitioned as finely as the block level.
As a result of double-counted or erroneously enumerated persons, the actual
number of persons missed by the Census, the gross undercount, may be substantially
larger than the net undercount.
Even in the absence of any net population undercount, there still could
be a substantial differential undercounting of minorities. That is, a small
over counting of whites could be offsetting a relatively substantial under
counting of minorities. See, Eugene P. Ericksen, "Who Gets Missed in
the Census? Evaluating the Net and Differential Undercount," Dec. 2000,
A. White and K. Rust, eds., National Research Council, Preparing for
the 2000 Census: Interim Report II, 1997; U.S. Department of Commerce,
Bureau of the Census, Census 2000 Operational Plan, 1997; Allan J.
Lichtman and Samuel Issacharoff, "Adjusting Census Data for Reapportionment:
The Independent Role of the States," Journal Of Litigation (Dec.
Their website is >www.ncec.org<.
In a 1999 decision, the United States Supreme Court ruled that apportionment
of congressional districts among states for the post-2000 redistricting
must be based on uncorrected census numbers. However, this was a statutory,
not a constitutional interpretation. Congress could have changed the rules
for 1990 and may still do so at a future time.
The greater the minority population of a district, the greater, in general,
the opportunities for minorities to participate fully in the political process
and elect candidates of their choice. This relationship applies even if
no single minority group has a population or voting-age majority in the
district. The vast majority of minority group members elected to Congress
and state legislatures are elected in majority-minority districts. For diverse
states, the analysis will focus on the overall minority percentage. It is
beyond the scope of the study to scrutinize separately each minority group
within each district. However, it is likely that the use of corrected census
data will most significantly enhance the representation of the largest minority
group within the district.
With 53 congressional districts and a corrected population of 30,597,578
for California, the ideal congressional district would have consisted of
577, 313 persons, only about 5,000 persons higher than the population of
the ideal congressional district with uncorrected data and 52 districts.
The population of an ideal district for statewide congressional or state
legislative plans equals the state's total population divided by the number
of districts statewide.
For states required by Section 5 of the Voting Rights Voting Act to pre-clear
redistricting plans with the United States Department of Justice, Courts
have mandated that the Department assess whether a new plan results in the
retrogression of minority voter opportunities. For states not covered under
Section 5, retrogression could result in a violation of Section 2 of the
Voting Rights Act, which applies to all states and localities.
The corrected minority percentage for this and other districts analyzed
in this report is obtained by aggregating the corrected data, partitioned
by race, for the census blocks included in the district.
One-person-one-vote requirements are much less stringent for state legislative
than for congressional plans. A state legislative plan presumptively meets
one-person-one-vote requirements if its total deviation from the population
of the ideal district is 10 percentage points or less. Thus a plan would
be presumptively valid if its least populated district was 5 percent below
the ideal population and its most populated district was 5 percent above
the ideal population.