The Poll of Polls

As part of TIME'S '72 election-year coverage, Senior Correspondent John Steele will assess the performance of professional pollsters from time to time. The New Hampshire primary was their first test. Steele's report:

EVEN expert pollsters shudder when they contemplate primary elections. The voting base is small, the electorate volatile, the reins of party discipline lax--and in 1972 the Democratic candidates many. Pollster Louis Harris, who is not gauging primaries this year, points out that polling must continue virtually up to election eve to spot possible switches in voter sentiment that can run as high as 20% in the week before voting day. Given all the hazards, the polling in New Hampshire measured up reasonably well against the actual results.

The most thorough polls were taken for the Boston Globe by Becker Research Corp. of Boston, using sophisticated but by no means risk-free telephone polling. Becker's first poll in January showed Edmund Muskie receiving 65% of the Democratic vote, a figure that soon became a target in election stories and drew cries from the Muskie camp that their candidate was running against a phantom, i.e., the numbers.

What was to have been the final poll was taken in late February and showed a drastic change. There had been a massive erosion of Muskie strength to 49%, and a major growth in George McGovern strength of 13 points to 31%. Even as Becker's pollsters made their second round of 435 sampling calls, Muskie was mounting a flatbed truck in the campaign's single dramatic moment to denounce, in tears, Manchester Publisher William Loeb. Another "final" poll was ordered by the Globe; it showed Muskie strength off still more to 42%, McGovern at 26%, and the number of undecideds inexplicably doubled to 20%.

Actually, the second final poll was rushed to meet a deadline, and depth questions were not included. The first final, as a result, came closer to the actual outcome. It was only one point off Muskie's 48% vote total, off six on McGovern's 37%, two points off on Wilbur Mills and a point off each on Sam Yorty and Vance Hartke.

Two other polls, meanwhile, had not done as well. The Public Television Service in New Hampshire took an early February sample that showed Muskie receiving 58% of the vote and McGovern 19%; it was never updated. Pat Caddell, the 22-year-old president of Cambridge Survey Research, in private polls for McGovern, gave Muskie 46% and McGovern 30%. Caddell did, however, advise correctly that McGovern could do well in blue-collar and middle-income areas by taking specific though not necessarily popular stands on issues. Overall, the polling in the nation's first primary was fairly accurate.

This file is automatically generated by a robot program, so reader's discretion is required.