Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 1 of 17 EXHIBIT C
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 2 of 17
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 3 of 17 Response to Professor Alford s Supplemental and Rebuttal Reports I. Statement of Inquiry 1. I have been asked to examine two supplemental reports filed by Professor John Alford, one filed on March 14, 2014 (supplemental), and one on April 14, 2014 (rebuttal). In this regard, I have been asked to examine voting patterns and demographic data in congressional districts in the State of Texas. I have been retained by the Rodriguez plaintiffs in this case. I am paid $400 an hour, which is my standard consulting rate for such cases. II. Background and Qualifications 2. My background and qualifications are presented in my initial report in this case, dated February 26, 2014. III. Sources of Information 3. Data on population and voting in the Voting Tabulation Districts (VTD) and Congressional Districts (CD) come from the redistricting website of the Texas Legislative Council. (http://www.tlc.state.tx.us/redist/redist.html) 4. Data on Citizen Voting Age Population (CVAP) come from the American Community Survey (ACS) and from the 2000 Census. Specifically, I examined the 1
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 4 of 17 2000 Census estimates, based on the long form of the Census in that year, the 5-year average of the ACS from 2005-2009 for estimates of the CVAP for 2007 and the 5- year average of the ACS from 2008-2012 for estimates of the CVAP for 2010. I rely on three reports produced by the Texas Legislative Council for estimates of CVAP in CDs under the district map in force from 2006 to 2010 (Plan C100). The 2000 Census estimates are presented in RED-201, Expanded District Analysis with Citizenship, Congressional Districts PlanC100. Referred to below as RED-201. The 2008-2012 ACS data are presented in RED-116, American Community Survey Special Tabulation Using Census and American Community Survey Data, Congressional Districts PlanC100. Referred to below as RED-116. 5. I reference several other reports here that I have filed in other cases relating to redistricting in the State of Texas. These are Report on Minority and White Representation and Voting Patterns in the Texas Congressional District Plan C185, dated August 8, 2011, in Perez v. Perry, U.S. District Court of the Western District of Texas; Rebuttal Report to the Supplemental Expert Report of Professor John Alford, dated January 16, 2012, State of Texas v. United States, U. S. District Court of the District of Columbia; Analysis of Racial Voting Patterns in the 2004 Democratic Primary in Congressional District 25, Llyod Doggett v. Leticia Hinojosa, Dated October 27, 2011, Perez v. Perry, U.S. District Court of the Western District of Texas; and Analysis of Racial Voting Patterns in 2012 Democratic Primary in Texas Congressional District 35, Lloyd Doggett versus Sylvia Romo and Maria Luisa 2
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 5 of 17 Alvarado, June 1, 2012, State of Texas v. United States, U. S. District Court of the District of Columbia. IV. Findings A. Voting Patterns in Congressional District 25 (CD 25) 6. On Page 23 of his March 14, 2014, Report, Professor Alford claims that voters in Travis County are more polarized than in any other part of the state. No factual reference is offered for this statement in the report. 7. My reports in Perez v. Perry and United States v. State of Texas show that in general elections Travis County has the lowest racial polarization of the most populous counties in the state because a high percentage (but still less than half) of Whites vote for the candidates preferred by majorities of Hispanics and Blacks. That is, in general elections in Travis County majorities of Blacks and of Hispanics regularly vote for the same candidates and a substantial fraction of Whites crossover and support those candidates as well. Specifically, see Table 5 on page 55 of my Report on Minority and White Representation and Voting Patterns in the Texas Congressional District Plan C185, dated August 8, 2011; pages 5-6 and 15-24 of my Rebuttal Report to the Supplemental Expert Report of Professor John Alford, dated January 16, 2012; and my Analysis of Racial Voting Patterns in 2012 3
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 6 of 17 Democratic Primary in Texas Congressional District 35, Lloyd Doggett versus Sylvia Romo and Maria Luisa Alvarado, June 1, 2012. 8. Professor Alford states that the distinctive feature of Travis County electorate is that there is a high fraction of Democrats among whites. Professor Alford offers no evidence or analysis that he used in making this statement, beyond voting behavior in general elections. No analysis of party registration or party identification of Whites in this county is offered. No numerical estimate is offered of the fraction of Democrats who are White in Travis County. 9. Professor Alford states on page 23 of his report that it is his conclusion that Blacks and Hispanics are not cohesive in primary election voting in Travis County. My prior assessment is that Blacks and Hispanics exhibit a high rate of cohesion because in most primary elections large majorities of these two groups vote for the same candidate, and the candidates preferred by minorities won 75 percent of the time in Travis County primaries. See for example pages 24-25 of my Rebuttal Report to the Supplemental Expert Report of Professor John Alford, dated January 16, 2012. The primary election process, then, in Travis County and under old CD25 did not prevent minority-preferred candidates from reaching the general election contests. 10. In the two endogenous Congressional primary elections examined in prior reports, the minority preferred candidate won the primary. See my Analysis of 4
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 7 of 17 Racial Voting Patterns in the 2004 Democratic Primary in Congressional District 25, Lloyd Doggett v. Leticia Hinojosa, Dated October 27, 2011, and Analysis of Racial Voting Patterns in 2012 Democratic Primary in Texas Congressional District 35, Lloyd Doggett versus Sylvia Romo and Maria Luisa Alvarado, June 1, 2012. 11. Hence, it is my conclusion that CD 25 under Plan C100 was a district in which Hispanics and Blacks had the ability to elect their preferred candidates. Primary elections in Travis County and in CD 25 under PlanC100 did not prevent minoritypreferred candidates from winning nomination to the general election. B. Growth in Citizen Voting Age Population 12. Professor Alford examines growth in Hispanic adult citizen population in specific CDs as well as the Dallas-Forth Worth and Houston areas in his report Rebuttal Report of John R. Alford, Ph.D., Perez v. Perry (W.D. Tex.), dated April 14, 2014. 13. Professor Alford makes three arguments. First, Hispanic CVAP growth occurred everywhere throughout the state. This, he suggests, makes it difficult to draw new minority districts. Second, the areas where there was concentrated growth of Hispanics either (i) were already majority HCVAP CDs, so nothing could be done to alter their representation, or (ii) did not see corresponding growth in HCVAP, so it is difficult to create CDs in which minorities had the ability to elect their preferred candidates. Third, the hypothetical or potential gains of minorities are achieved in 5
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 8 of 17 alternative maps only by making districts more irregular. However, no measures of district compactness or irregularity of boundaries are offered. 14. Turning to his first argument, it is indeed the case that Hispanic population and Hispanic CVAP increased everywhere in the state. Table 2 in Professor Alford s report shows that every CD in Plan C100 experienced increases in absolute numbers of HCVAP and in the percentages of CVAP that are Hispanic over the decade. The CDs in Plan C100 with the largest increase in absolute numbers of HCVAP from 2000 to 2010 are CDs 23, 28, 16, 15, 21, 27, 10, and 22. HCVAP in each of these CDs increased by at least 50,000 persons, ranging from an increase of 53,929 in CD 22 to 88,364 in CD 23. 15. The change in the racial composition of districts was not uniform across the state. HCVAP is only part of the overall picture. For completeness I present Anglo CVAP (ACVAP) and Black CVAP (BCVAP) by district. Tables 1 and 2 presents the numbers and percentages of ACVAP and BCVAP in 2000 and 2010 in CDs under Plan C100. Estimates of the ACVAP and BCVAP are estimated using Red-201 and Red- 116 by multiplying the total CVAP times the Percent Anglo of the CVAP and the Percent Black of the CVAP. Because the percentages are given to one decimal, a small rounding error arises with the estimates. These tables supplement Table 2 in Professor Alford s report. 6
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 9 of 17 16. On average, the HCVAP increased by 4.4% across all districts. On average, the ACVAP decreased by 6.5% across all districts. On average BCVAP increased 1.9%. 17. The Black CVAP increased as a percent of the population in 20 CDs, decreased in 11 CDs, and was unchanged in 1 CD. The CDs with the greatest increase in Black CVAP as a percent of total CVAP are CDs 6, 7, 22, and 24. See Table 1. 18. The Anglo CVAP decreased as a percent of the population in every district under Plan C100, and in half of the districts declined in absolute numbers. See Table 2. 19. Table 3 presents the districts ordered from those with the greatest to those with least decline in ACVAP as a percent of CVAP from 2000 to 2010. The CD with the least decline in ACVAP relative to CVAP was CD 25. That is, in this district growth in minority adult citizen population roughly kept pace with growth in Anglo adult citizen population. CDs 23, 26, 8, 28 round out the list of the CDs with the lowest decline in ACVAP as a percent of CVAP. 20. The CDs with the largest decline in ACVAP as a percent of CVAP were CDs 22, 24, 7, 29, 2, 3, and 10. All had declines of at least 10 percent. CD 22, south and west of Harris County, had the greatest decline in ACVAP%, dropping from 68.0% to 53.3% in Plan C100. Under Plan C235, the percent Anglo CVAP is increased to 54.3%. 7
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 10 of 17 21. Professor Alford states that the growth in Hispanic population occurred in CDs that were already majority Hispanic. He focuses on CD 29 and Harris County, as an example. CD 29 had the highest rate of growth in HCVAP as a percent of population of all CDs, rising from 46.8% in 2000 to 59.0% in 2010. As discussed below, the other majority Hispanic CDs in Plan C100 were in the South and West Texas area and the growth in Population and HCVAP was sufficient to create another majority HCVAP district in that area without disrupting other nearby districts in which minorities had the ability to elect their preferred candidates. As for CD 29, even though it had grown to 59.0% HCVAP by 2010, redistricting increased the percent HCVAP further, to 59.8% under Plan C235. 22. In the Houston area, CDs 22, 7, 10 and 2 under Plan C100 witnessed substantial declines in ACVAP as a percent of CVAP from 2000 to 2010. Notably, CD 22 had a very large increase in minority population and was on the cusp of becoming majority minority in 2010. The ACVAP as a percent of CVAP fell from 68.0% to 53.3% in CD 22 over the decade. In other words, redrawing this district s boundaries in order to reduce ACVAP by just 3.3 percent of ACVAP would convert this into a district in which minorities were a majority of the CVAP. This is the majority Anglo CD in Plan C100 that was closest to becoming majority minority. 23. Plan C235 increased the Anglo CVAP of CD 22 by 1%. The reconfiguration of CD 22 under Plan C235 decreased the Hispanic CVAP by 0.9%, decreased the Black CVAP by 2.1%, and increased the Asian CVAP by 2.0%. 8
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 11 of 17 24. The central thrust of my prior report in this case was to show that there was the potential to create an additional majority HCVAP CD in South and West Texas without disrupting districts in which minorities have the ability to elect their preferred candidates elsewhere in the State. My previous report in this case demonstrated that there was sufficient population in the VTDs in the majority Hispanic CDs (using all VTDs put in majority HCVAP CDs under either C100 or C235) in this region to create an additional CD without disrupting districts elsewhere in the state. The statistics that Professor Alford offered in his Table 2 and those in Tables 1 and 2 here reinforce that conclusion. 25. Table 4 presents the average growth in CVAP, HCVAP, ACVAP, and BCVAP in the 6 CDs in South and West Texas under Plan C100 and in CDs elsewhere in the State of Texas. Under Plan C100 these are CDs 15, 16, 20, 23, 27, and 28. I use the data in RED-201 and RED-116 for Plan C100 to calculated the racial compositions of the districts. 26. Total CVAP increased faster in CDs 15, 16, 20, 23, 27, and 28 under Plan C100 than in the rest of the state. The average growth in CVAP among these six CDs was 77,557 adult citizens. The average growth in CVAP among the remaining 26 CDs in the state under Plan C100 was 74,446 adult citizens. 9
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 12 of 17 27. Hispanics account for almost all of the growth in CVAP in CDs 15, 16, 20, 23, 27, and 28 under Plan C100, and Hispanics were the single largest group contributing to growth in CVAP throughout the rest of the state. Among the six CDs in South and West Texas under Plan C100, 87.1 percent of the growth in CVAP was due to growth in HCVAP (i.e., 67,545/77,557). Among the remaining 26 CDs in Texas, 42.3 percent of the growth in CVAP was due to growth in HCVAP (i.e., 31,504/74,446), 24.8 percent of the growth in CVAP was due to growth in ACVAP (i.e., 18,459/74,446), and 20.3 percent of the growth in CVAP was due to growth in BCVAP (i.e., 15,137/74,446). 28. Plan C235 removed a substantial Hispanic population from this set of districts in the Nueces County area in its configuration of what is the new CD 27. CD 34 under Plan C235 contains most of the population from Plan C100 s CD 27. A seventh majority Hispanic CD, CD 35 in Plan C235, is constructed in the area from Bexar County to Travis County. Doing so requires taking apart CD 25 in Plan C100, which was a district in which minorities could and did elect their preferred candidates. 29. I offered Plan C220 in my earlier report to show that it is possible to construct an additional CD in South and West Texas that is majority HCVAP without disrupting other nearby districts in which minorities have the ability to elect their preferred candidates. Specifically, Plan C220 constructed an additional majority HCVAP district (CD 35 in Plan C220); it kept CD 27 under Plan C100 largely intact, and it 10
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 13 of 17 retained a CD in the Travis County area (CD 25) in which minorities have the ability to elect their preferred candidates. 30. Plan C220 provides an example of a plan drawn in accordance with traditional districting principles that increases the number of CDs in which minorities have the ability to elect their preferred candidates. Other district configurations and maps might achieve further gains. My referencing Plan C220 is not meant to preempt other possible maps but to demonstrate that it was possible to draw at least one more majority Hispanic CD in the State of Texas without disrupting existing districts in which minorities have the ability to elect their preferred candidates. 11
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 14 of 17 Table 1. Black Citizen Voting Age Population 2000 to 2010 under Plan C100 District 2000 BCVAP 2010 BCVAP BCVAP 2000 BCVAP% 2010 BCVAP% BCVAP% 1 83,877 75,920-7,957 18.2% 73.4% 0.5% 2 85,957 119,927 33,971 19.3% 59.2% 3.6% 3 39,296 60,348 21,053 9.9% 64.8% 2.0% 4 47,262 62,488 15,226 10.3% 79.7% 0.4% 5 54,640 70,407 15,767 12.2% 71.5% 2.3% 6 54,802 90,065 35,264 12.9% 65.4% 4.1% 7 25,926 53,976 28,050 6.1% 63.7% 4.7% 8 40,799 39,821-978 8.9% 81.5% -0.6% 9 156,141 173,235 17,095 46.7% 18.2% 2.1% 10 40,967 75,144 34,177 9.6% 64.4% 2.7% 11 19,338 20,789 1,451 4.3% 66.7% -0.1% 12 25,816 39,067 13,252 6.1% 71.9% 1.2% 13 27,107 29,542 2,435 5.9% 75.8% 0.4% 14 45,030 53,451 8,420 10.2% 64.9% -0.1% 15 9,908 9,497-411 2.7% 22.4% -0.5% 16 14,693 15,966 1,273 4.1% 19.2% -0.4% 17 46,055 53,244 7,189 10.0% 10.0% 0.0% 18 184,665 189,905 5,241 48.1% 46.7% -1.4% 19 24,697 27,214 2,517 5.4% 5.6% -0.2% 20 32,353 37,477 5,125 7.7% 8.2% 0.5% 21 28,978 40,207 11,229 6.2% 6.5% 0.3% 22 41,386 87,204 45,819 9.9% 14.9% 5.0% 23 13,903 19,110 5,207 3.5% 3.6% 0.1% 24 41,169 77,949 36,780 9.9% 15.7% 5.8% 25 45,864 46,222 357 10.5% 8.6% -1.9% 26 68,169 83,711 15,542 15.7% 14.3% -1.4% 27 11,959 11,400-559 3.0% 2.5% -0.5% 28 5,882 8,467 2,585 1.7% 1.9% 0.2% 29 41,864 46,676 4,811 13.9% 15.1% 1.2% 30 186,288 212,963 26,675 50.4% 52.2% 1.8% 31 58,159 77,912 19,753 13.1% 12.7% -0.4% 32 33,639 34,052 413 9.1% 9.5% 0.4% Source: Calculated by author from Red-201 and Red-116 for Plan C100. 12
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 15 of 17 Table 2. Anglo Citizen Voting Age Population 2000 to 2010 under Plan C100 District 2000 ACVAP 2010 ACVAP ACVAP 2000 ACVAP% 2010 ACVAP% ACVAP% 1 353,020 297,997-55,024 76.6% 73.4% -3.2% 2 311,314 310,030-1,284 69.9% 59.2% -10.7% 3 299,283 328,620 29,337 75.4% 64.8% -10.6% 4 384,518 465,448 80,930 83.8% 79.7% -4.1% 5 352,027 347,179-4,848 78.6% 71.5% -7.1% 6 319,039 346,496 27,447 75.1% 65.4% -9.7% 7 329,384 318,357-11,028 77.5% 63.7% -13.8% 8 385,526 391,017 5,490 84.1% 81.5% -2.6% 9 91,611 64,608-27,003 27.4% 18.2% -9.2% 10 317,921 393,439 75,518 74.5% 64.4% -10.1% 11 326,046 330,155 4,109 72.5% 66.7% -5.8% 12 326,291 384,784 58,493 77.1% 71.9% -5.2% 13 366,639 355,441-11,197 79.8% 75.8% -4.0% 14 304,176 343,461 39,285 68.9% 64.9% -4.0% 15 100,915 96,701-4,214 27.5% 22.4% -5.1% 16 87,084 82,851-4,233 24.3% 19.2% -5.1% 17 361,069 400,391 39,322 78.4% 75.2% -3.2% 18 110,569 99,629-10,939 28.8% 24.5% -4.3% 19 321,518 322,198 680 70.3% 66.3% -4.0% 20 123,529 113,346-10,183 29.4% 24.8% -4.6% 21 342,126 411,964 69,838 73.2% 66.6% -6.6% 22 284,266 311,946 27,680 68.0% 53.3% -14.7% 23 148,168 182,075 33,907 37.3% 34.3% -3.0% 24 308,561 295,908-12,653 74.2% 59.6% -14.6% 25 276,060 333,225 57,165 63.2% 62.0% -1.2% 26 319,569 413,871 94,302 73.6% 70.7% -2.9% 27 140,318 137,250-3,068 35.2% 30.1% -5.1% 28 102,071 120,770 18,698 29.5% 27.1% -2.4% 29 112,341 74,186-38,155 37.3% 24.0% -13.3% 30 120,496 104,034-16,462 32.6% 25.5% -7.1% 31 318,319 411,032 92,712 71.7% 67.0% -4.7% 32 262,825 229,043-33,782 71.1% 63.9% -7.2% Source: Calculated by author from Red-201 and Red-116 for Plan C100. 13
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 16 of 17 Table 3. Anglo Citizen Voting Age Population 2000 to 2010 under Plan C100, CDs ordered from greatest to least change in ACVAP% District 2000 ACVAP 2010 ACVAP ACVAP 2000 ACVAP% 2010 ACVAP% ACVAP% 22 284,266 311,946 27,680 68.0% 53.3% -14.7% 24 308,561 295,908-12,653 74.2% 59.6% -14.6% 7 329,384 318,357-11,028 77.5% 63.7% -13.8% 29 112,341 74,186-38,155 37.3% 24.0% -13.3% 2 311,314 310,030-1,284 69.9% 59.2% -10.7% 3 299,283 328,620 29,337 75.4% 64.8% -10.6% 10 317,921 393,439 75,518 74.5% 64.4% -10.1% 6 319,039 346,496 27,447 75.1% 65.4% -9.7% 9 91,611 64,608-27,003 27.4% 18.2% -9.2% 32 262,825 229,043-33,782 71.1% 63.9% -7.2% 5 352,027 347,179-4,848 78.6% 71.5% -7.1% 30 120,496 104,034-16,462 32.6% 25.5% -7.1% 21 342,126 411,964 69,838 73.2% 66.6% -6.6% 11 326,046 330,155 4,109 72.5% 66.7% -5.8% 12 326,291 384,784 58,493 77.1% 71.9% -5.2% 15 100,915 96,701-4,214 27.5% 22.4% -5.1% 16 87,084 82,851-4,233 24.3% 19.2% -5.1% 27 140,318 137,250-3,068 35.2% 30.1% -5.1% 31 318,319 411,032 92,712 71.7% 67.0% -4.7% 20 123,529 113,346-10,183 29.4% 24.8% -4.6% 18 110,569 99,629-10,939 28.8% 24.5% -4.3% 4 384,518 465,448 80,930 83.8% 79.7% -4.1% 13 366,639 355,441-11,197 79.8% 75.8% -4.0% 14 304,176 343,461 39,285 68.9% 64.9% -4.0% 19 321,518 322,198 680 70.3% 66.3% -4.0% 1 353,020 297,997-55,024 76.6% 73.4% -3.2% 17 361,069 400,391 39,322 78.4% 75.2% -3.2% 23 148,168 182,075 33,907 37.3% 34.3% -3.0% 26 319,569 413,871 94,302 73.6% 70.7% -2.9% 8 385,526 391,017 5,490 84.1% 81.5% -2.6% 28 102,071 120,770 18,698 29.5% 27.1% -2.4% 25 276,060 333,225 57,165 63.2% 62.0% -1.2% Source: Calculated by author from Red-201 and Red-116 for Plan C100. 14
Case 5:11-cv-00360-OLG-JES-XR Document 1065-3 Filed 06/09/14 Page 17 of 17 Table 4. in Total, Hispanic, Anglo, and Black CVAP from 2000 to 2010, Comparison of South and West Texas Majority Hispanic CDs (CDs 15, 16, 20, 23, 27, and 28) versus All Other CDs in the State of Texas Districts SW Texas Majority Hispanic (15, 16, 20, 23, 27, 28) Average CD Total CVAP Average CD Hispanic CVAP Average CD Anglo CVAP Average CD Black CVAP Number of CDs 77,557 67,545 5,151 2,203 6 All other CDs 74,446 31,504 18,459 15,137 26 15