San Francisco Chronicle

Friday, June 5, 1998

Prop. 227 Got Few Latino Votes
Early polls had claimed more minority support
By RAMON G. McLEOD, MARIA ALICIA GAURA, Chronicle Staff Writers

Contrary to public opinion polls that had shown Latinos in favor of Proposition 227, precinct results and statewide election exit polls show that Latinos voted heavily against the bans on bilingual education.

Backers of Proposition 227, which won with a commanding 61 percent of the votes on Tuesday, had pointed with some pride to what appeared to be support for the measure that cut across racial lines.

But although whites and Asian Americans largely supported it, most African Americans and Latinos -- the biggest non-English speaking group in California -- resoundingly voted ``no.''

Latino support, which some polls found to be as high as 61 percent two months ago, had evaporated by election day, according to exit polls conducted by CNN and the Los Angeles Times.

``I think what happened was that people finally understood that this was about doing away with bilingual education, not correcting it,'' said Ortencia Lopez, executive director of El Concilio of San Mateo, an umbrella organization for Latino nonprofit groups. ``Many of us agree that bilingual education needs fixing, but this goes far beyond that.''

In the 1990 Census, at least one third of Latinos reported that they did not speak English very well or at all, and demographers believe those rates probably are about the same today.

Asian Americans tend to be more proficient in English, with only about a quarter reporting poor English skills. Also, although exit polls showed that 57 percent of Asian American voters approved the measure, in places where non-English-speaking Asian Americans are highly concentrated, they too rejected 227.

Exit polls in San Francisco showed that 74 percent of Asian Americans voted against the measure, similar to San Jose precincts with large, non-English-speaking Asian American populations.

``I think you can say pretty clearly that groups that were most directly affected by this were very much against it, both Latinos and, here in the city, Asians,'' said Tom Hsieh Jr., a pollster who conducted Tuesday's exit polls of San Francisco's Asian American voters.

Ron Unz, the millionaire Silicon Valley businessman who authored the measure, said that these kinds of exit polls are unreliable indicators.

``I think you need to be very, very skeptical about these kinds of polls,'' he said. ``First of all, you are polling only people willing to stop and talk, and there may have been some peer pressure on them to say what they thought needed to be said.

``If you took a poll at the public forums we did, you would have thought only 5 percent were in favor of 227.''

Harry Patron, president of the Tomas Rivera Policy Institute, a Southern California think tank that studies Latino issues, agreed with Unz about flaws in polling, but for a very different reason. ``We think the polls have basic problems when it comes to Latinos,'' Patron said. ``It happened with Propositions 187 and 209 too.''

Patron said there are three reasons why pollsters seem to misinterpret Latino voting trends.

``First, even though Latinos comprise one of every three or four state residents, they make up only 13 to 14 percent of the voting pool,'' Patron said. As a result, polls include few Latinos, and the margin of error is large.

Also, many Latino voters are predominantly Spanish-speaking, and Patron thinks pollsters are ill-equipped with bilingual staff.

``I suspect that they are overwhelmed with the Spanish speakers and so seek out the English-speaking Latino voters,'' Patron said.''

Furthermore, Patron said that working class and poor Latinos tend to ignore elections until the last weeks of the campaign and then rely heavily on the influence of opinion-makers when deciding how to vote.

``You see a tremendous fluidity in the two weeks preceding an election,'' Patron said. ``They take their cues from political leaders, and those cues are often not forthcoming until the end.

``Those factors contribute to the discrepancy between polling and voting, which we as an institute have been trying to point out for years.''