An overview of this methodology was first presented in the 6th Annual Conference of the DANR, “Dominican American Future: Health & Economic Empowerment. This presentation provided an explanation of the data obtained, preliminary findings and data limitations. For a brief discussion of terms used and reasoning of approach regarding this study, please refer to Appendix 1. This section discusses how the congressional district population data was used to extract the number of adult citizens. Appendix 2 is a listing of the top 100 Dominican populated Congressional Districts (CD) in the 108th Congress. This table compares, at a CD level, Census based Dominican population estimates with the respective population count of the other five major Latino ethnic groups in the US (including Puerto Ricans). Appendix 3 is a list of the top 100 congressional districts with the most adult Dominican citizens. Appendix 4 is a chart of the state- and group-specific adult citizenship rates; these figures were used at the congressional district levels to obtain the table illustrated in Appendix 3. Estimating Adult Citizens at a Congressional District Level To our knowledge, the US Census is the only publicly available data source that provides Congressional District level population estimates. This data was prepared using the official numbers in the 2000 Census. Technically, these numbers were drawn from the short form of the 2000 Census information, and then distributed by the congressional districts using the maps created before the election of 2002, after the decennial constitutional redistricting process. However, the first and yet only official, US Census 2000 data does not account for the historical undercount of minorities and the new problem of “misrepresentation” of some Hispanic groups acknowledged later by the US Census Bureau.55 Although, an attempt was made to correct the data, the result was not sufficiently reliable. When compared to other data source options, the improvement of the data was minimal. 20 Given that the number of citizens by congressional districts is unknown, a variation of the shift-share technique,56 was employed as the method for estimating this value. The Census’ Summary File 4 was used to obtain the number of citizens, adults and adult citizens at the state level for each of the six major Latino groups (data was available for almost all states).57 The respective percentages were then calculated based on the base population for each ethnic group and applied to statewide 100-Percent data percentages for each Hispanic ethnic group at the congressional district level. 58 Thus, it is assumed that the demographic age and the citizenship distribution at the state level is proportional to that of the group’s distribution at the congressional district level. While this holds true for states that have only one Congressional District (e.g. Alaska, Delaware, Puerto Rico, and Washington DC), this ratio does not necessarily always hold true in states with multiple Congressional Districts, like New York. Box 1 illustrates the method used. While data has been gathered for all congressional districts in the country, a major part of the research effort focused on those states with congressional districts with over 500 Dominican constituents.

More specifically, the population calculations of the Pew Hispanic Center was used as a low range estimate, the revised population count done by the US Census as a mid-range estimate, and the SUNY-Albany/Lewis Mumford Center’s numbers as a high-end estimate. Given these figures, the following logarithmic formula was applied to solve 22 for the annual population growth rate: P2000 = P1990 etr, where P2000 is the estimated population in 2000, P1990 is the estimated population in 1990, t is time (in years), and r is the annual rate of population growth. The 1990 base population is varied by population estimates. The derived range of annual growth rates was used for each population scenario. Finally, each scenario is varied through time from 2004 to 2020 in consecutive two-year increments. Thus, the three-dimensional potential population projection matrix provided a count range for the Dominican community that presently resides in the US. While the logarithmic population formula is a good baseline estimate for the Dominican community, it fails to account for net migration. In fact, net migration is often the most difficult and uncertain aspect of population projections even within the US Census figures.62 As such, based on the Bureau of Citizenship and Immigration Services’ (formerly known as the INS) entry chart, the percentage of Dominicans granted entry to the US, out of the entire immigrant population, was calculated. However, it was not possible to project future net immigration numbers, assuming current entry patterns. Moreover, immigration rates, both documented and undocumented are highly unpredictable since it often depends on the economic situation observed in the Dominican Republic at that time. These net migration figures, if calculated with some certainty, should ultimately be added to the residential population estimates from 2004 to 2020 to obtain final population projections. The goal of these population projections is to then estimate what percentage of the Dominican population is ultimately going to be eligible to vote nationwide. A look to the increasing naturalization rates provides a first line estimate. A conservative approach is used to estimate adult citizenship rates from 2004 to 2020. Summary The two main methodologies used were the shift-share projections to estimate the percentage of adult citizens per Latino group and the simple annual growth logarithmic formula to estimate the Dominican population from 2004 to 2020.