[EM] Debate Over Measuring Voter Turnout

[DEMOREP1 at aol.com]

Debate Over Measuring Voter Turnout

By GENARO C. ARMAS
  
WASHINGTON (AP) - Voters must be citizens, yet election turnout and 
registration rates typically have been measured using all voting-age 
residents, citizen or not. 

With increased immigration in recent decades, critics argue that removing 
noncitizens among the potential voter pool yields a turnout rate more 
encouraging than previously reported. 

For instance, results released Wednesday from a Census Bureau survey taken 
two weeks after the November 2000 election showed that 64 percent of U.S. 
residents at least 18 years old were registered to vote, and 55 percent 
voted. 

The proportions are significantly higher, however, when looking only at 
citizens of voting age who were surveyed - 70 percent registered, 60 percent 
voted. The report was the bureau's most detailed look at voting habits of 
citizens. 

Making that distinction shows the voter turnout decline is not as serious as 
feared, said Michael McDonald, a political science professor at the 
University of Illinois in Springfield, Ill. 

``It makes a huge difference in how we develop strategies to increase turnout 
rates and the perception of people's trust in government,'' McDonald said 
Tuesday. 

He estimated that 10 percent of the voting-age population in 2000 could not 
vote because they lacked U.S. citizenship or were in prison, up from 2 
percent in 1972. 

The census report said turnout for the presidential election increased from 
54 percent to 58 percent in 1996 when noncitizens were removed from the 
measurement. Similarly, the turnout increased from 61 percent to 68 percent 
in 1992, from 57 percent to 62 percent in 1988, from 60 percent to 65 percent 
in 1984 and from 60 percent to 64 percent in 1980. 

There are drawbacks to looking only at voting-age citizens, bureau analyst 
Jennifer Day said. Most importantly, research into the topic is less accurate 
before 1994, making comparable long-term comparisons difficult, she said. 

That is why the bureau still will measure turnout and registration rates for 
all voting-age American residents as well, she said. 

The survey covered just the noninstitutional population, so it did not 
account for people in prison, for instance. In most states, convicted felons 
cannot vote. 

The survey typically has overstated both registration and turnout. The survey 
estimated 111 million votes were cast in the 2000 election, but the official 
count was less than 106 million votes. 

The discrepancy is due in part to some people claiming on the survey to have 
registered and voted when they did not. 

Election researcher Curtis Gans said the most accurate measure of turnout 
would account for many additional factors, including Americans who live 
abroad. 

Gans is director of the nonpartisan research group, the Committee for the 
Study of the American Electorate. He has estimated a 2000 turnout rate of 
just over 50 percent, up from 49 percent in 1996. 

But, Gans said, there remains a decline in the percentage of young people 
voting and an overall decline, though not as steep, in overall voter 
participation ``by any standard you want to use.'' 

Other findings from Wednesday's report: 

The discrepancy between the newer and older methods of calculating turnout 
was most evident in high immigration states, and among Hispanics and Asians, 
Day said. 

The percent of the voting-age population registered to vote in the 2000 
election was at an all-time low of 64 percent. Still, the 2000 election had a 
record number of people registered to vote, 130 million.