Technically a study week that represents the midway point through the semester, this break seems to be unique to Vassar. Most people I talk to from other schools drop their jaws in awe when they learn that I have an entire week free of classes every October. (Not to mention more than a month off from classes for the Christmas season.) True, Vassar may give us more time off from classes than most other schools give their students, but I'm sure the amount of work we have no doubt makes up for it.

This break is an unusual one for me; it is the first since my freshman year that I've not flown home to Sacramento. Rather, I'm spending the first part of the week with a friend in beautiful Branford, Conn. (where I'm watching the leaves outside change colors as I edit this week's newsletter on my trusty laptop), then returning to campus for the rest of the week.

I'll likely use the time at campus before classes start again to work on the dreaded "real world" job hunt of the soon-to-be college graduate. Although for awhile I was considering going immediately into grad school, I've since decided to take a few years to travel and work before returning to school.

The first step of the process is to figure out what I want to do with my life.

Thanks for reading, and have a great week!

-Gabe

SACRAMENTO, Calif. -- If you've ever wondered why political reporters insist on writing about opinion polls during an election year, there is a wonderful scene in the movie "African Queen" that says it all.

Humphrey Bogart starts kicking his boat's ancient steam valve, and Katherine Heburn watches him, horrified. He has to kick it, he explains, to dislodge the rust that builds up every so often.

But why doesn't he just fix the valve? she asks, mystified.

"I kind of like kicking it. She's all I got," he answers.

And that, ladies and gentlemen, is precisely why political reporters write about polls: It's all they've got.

Barring scandal, political reporters during any given election year offer the public three things: profiles of the candidates, analyses of the television advertising and stories about the polls. But the polls invariably become the top stories, because they purport to reflect the deepest feelings of the electorate. Forget that they are often inaccurate, or that they imply a level of precision that they don't really have, or that they are subject to the vagaries of the pollsters' methodology. The polls play a significant role in an election because reporters give them coverage.

Even experienced reporters fall into the trap of falsely interpreting the accuracy of polls.

The race between Republican Matt Fong and Barbara Boxer is a perfect example. Since the spring, these two candidates have been locked in a tie in a November matchup. The candidates' own internal polling tells them this, so do the polls by a variety of organizations, including the Field Poll, the Mason-Dixon Poll and the Public Policy Institute of California.

In recent weeks, however, the polls have shown Fong with a slight edge. Four percentage points according to some, 2 percentage points according to another. And reporters have dutifully reported that Fong is ahead.

But the reality is far different. The most significant number in a survey about a close race isn't who is winning, it is the margin of error. For example, the margin of error in a recent survey in which Fong led 46 percent to 42 percent was plus or minus about 4 percent.

That means that anything up to 4 percent can be added or subtracted from each candidate's total and still accurately reflect the results of the survey. So Fong could be leading 50 percent to 38 percent, or the two could have been dead even at 46 percent, or Boxer could be leading 46 percent to Fong's 42 percent -- all are accurate products of the survey. The bottom line is that the race is too close to call, even though the poll's own numbers show the two candidates could be in a dead heat.

One Senate poll showed Fong leading by 2 percentage points, and reporters said the poll showed Fong was winning. But the margin of error was 3.5 percentage points, which means the race is a statistical tie.

Similarly, the polls in the gubernatorial race between Democrat Gray Davis and Republican Dan Lungren have shown Davis ahead by six or eight points, sometimes more. But when a margin of error of 3.5 percent is factored in, the contest is far closer, and in at least one case Lungren could be out in front.

The classic poll miscue was in 1982, when one statewide survey showed Los Angeles Mayor Tom Bradley victorious in the governor's race and Jerry Brown the winner of the U.S. Senate campaign. Both, of course, lost.

In the end, polls may show a general tendency of the public one way or another, but they are not precise indicators of the public's feelings. Moreover, the pollsters don't report how many people who were contacted refused to talk to the pollsters -- surely an important number. And on sensitive issues, such as race, the polls have often been found wanting. In 1982, for example, the question arose whether race would play a role in the election of Bradley, who is black. The polls, in effect, said the race issue was negligible, but afterwards experts believed those poll numbers were low because people were not inclined to admit that they would cast ballots on the basis of race. Similarly, exit polls in 1994 low-balled the turnout against Proposition 187, the anti-immigration initiative, perhaps because the depth of anti-Hispanic feeling was underestimated.

The lesson here is simple: All polls should be taken with a grain of salt.

Stay up-to-date on Legislative contact information. Order the newest edition of "the little red book" today! Place your secure online order through our Web site: http://www.capenq.com/order.

*** Legislative Wall Maps ***

Looking for a poster-size map of California Legislative districts, complete with Legislator photos around the border? Our Wall Map, now only $9.95, may be just what you're looking for! Check out http://www.capenq.com/brochure/wallmap.html for more information.

*** Secure Ordering ***

All '98 publications are available online. Place your SECURE online order today! Surf on over to http://www.capenq.com/order.