We know all about Clinton and Dole and the triumphant return of Congressman Bono, but we were curious about some of the other winners and losers in Election '96. So Outlook asked three writers to look at the implications for three enduring institutions: conservatives, liberals and Washington's Permanent Government of lobbyists and other Capitol Hill insiders.

ALL HAIL Zogby, the pollster who conquered the 1996 election.

And may you burn in the fires of polling hell, you lucky dog, hiss his competitors, who say John Zogby is the newest bad boy of survey research. They criticize some of his polling techniques as little more than methodological malpractice.

Well, in polling, getting it right is the best revenge. And that's exactly what Zogby did last week.

On Monday, Zogby predicted Bill Clinton would beat Bob Dole by 8.1 percentage points in the popular vote. He based his projection on the final result of daily tracking polls he had been conducting since summer for the Reuters news agency. Clinton won by 8.4 percentage points in unofficial returns. (The vote becomes official after each state certifies the results.)

"It felt great," said a weary but obviously pleased Zogby last week after a news conference held by Reuters at its Washington office.

Zogby felt particularly good because his polls had been widely criticized and widely ignored throughout much of the fall. They consistently showed Clinton with substantially narrower leads than other polls (including some Washington Post surveys that I directed). His results were so different -- sometimes off by more than 15 percentage points from other polls -- that they were dismissed by many news organizations.

His polling became easier to ignore when Zogby explained his unorthodox techniques, which defied some fundamental conventions of survey research.

Unlike other pollsters, Zogby doesn't randomly choose respondents from all households. Rather, he selects his sample from listed telephone numbers. Research suggests that surveys based on listed-number samples are far less accurate because people with un listed numbers are different from those who put their numbers in the phone book.

Zogby said those studies are no longer valid for today's population, which has more unlisted numbers. "When I see a recent study that shows otherwise, I'll change."

Nearly all other surveyors conduct interviews during evening hours, when more working people tend to be home. Zogby's interviewers call between 9 a.m. and 9 p.m., with about 30 percent of his sample contacted before 5 p.m. That produces a sample that contains more retirees and stay-at-home women.

No problem here either, he said -- he "weights" or adjusts down the number of women and older people to match census figures. And interviewing during the day allows him to interview people who work swing shifts -- voters, he said, who are missed by other pollsters. Zogby believes that most major media polls inadvertently over-represent Democrats (he's not alone in this criticism). To avoid this, Zogby weights his surveys by party identification so the adjusted sample is 34.5 percent Democrats, 34 percent Republicans and the rest independents.

This decision also raised eyebrows. Network exit polls on Tuesday found that 39 percent of voters called themselves Democrats and 35 percent identified themselves as Republicans. But even those figures are suspect because party identification rises and falls constantly, making any particular estimate unreliable. So where did his party numbers come from? He based them on past exit polls, surveys from other organizations -- and his gut feelings. "Polling is 80 percent science and 20 percent art," he said. Whether through science, art or just plain luck, Zogby's polling methods produced an accurate prediction -- making it hard, at least for the next four years, to ignore Zogby or his polls. "I started doing {polling} 13 years ago," Zogby said. "I know I was the new kid on the block. I'm not the new kid any more." Richard Morin is The Washington Post's director of polling.