Numbers you can count on

When former Alaska Gov. Sarah Palin visited Fort Bragg recently, some media outlets reported that 4,000 people attended her book signing.

The News & Observer did not use the 4,000 figure, which was supplied by the Army. In our article, we said more than 1,000 people were in line an hour before Palin was scheduled to arrive.

Some readers questioned why we didn't use the estimate of 4,000. We didn't use the larger figure because we didn't think it was accurate.

This was the latest story in which we struggled to reconcile our crowd estimate with an estimate from an organization or official source.

Martha Quillin, a 25-year veteran reporter, covered the Palin event for The N&O. She arrived at 10 a.m., an hour before Palin did.

The Army told her then that about 1,000 people were in line. Based on her observation, that seemed reasonable to Quillin. She went to the back of the line and spoke with the last person in line - Estimated Person No. 1,000.

Quillin saw Estimated Person No. 1,000 arrive at Palin's table 2.5 hours into the 3-hour event. That means Palin was greeting fans and signing books at the rate of about 6.7 people per minute. If Palin continued that rate for another 30 minutes, she would have greeted about 1,200 people.

The Army estimated another 500 to 700 were still in line when Palin left. Based on figures from the Army and our own observation, we concluded 1,500 to 2,000 people were in line to see Palin.

But when the event ended, the Army estimated 4,000 people had been in line.

This didn't seem consistent with the prior numbers supplied by the Army and with Quillin's observations.

For Palin to have greeted 3,500 people (the Army's estimate of 4,000 total minus 500 who didn't get to see her), Palin would have had to greet (and sign books) at a rate of 19.4 people per minute - or one every 3 seconds for three hours.

Fort Bragg spokesman Tom McCollum said he thinks that's about right - "She was flying on the books."

McCollum said he used a similar method to Quillin's to gauge how fast the line was moving; he watched a man with a big camera buy a book and then head toward the back of the line, though he didn't actually see where the man ended up. When the man reached Palin a little more than an hour later, McCollum estimated that Palin was seeing 1,100 to 1,200 people an hour.

We were skeptical of the Army's estimate of 4,000. So we left out that estimate and reported that more than 1,000 were in line an hour before Palin's scheduled arrival, which seemed beyond dispute.

A secondary headline said, "Hundreds greet former vice presidential candidate." That understated the crowd; we published a correction today on Page 2A.

When a gathering is politically charged, partisans on both sides use our crowd estimates as proof of our bias.

Our prominent coverage of Palin's recent visit generally was flattering to her. One of our harshest critics on the right wrote to praise our coverage; from the left, one woman wrote that our account of the Palin event was "worshipful and fawning."

Yet some Palin fans saw our not using the 4,000 figure as evidence that we were unfair to her.

When a reporter covers an event, estimating the crowd size is just one piece of the reporting - typically a small piece. Sometimes we take the official estimate with little scrutiny.

Parades are among the most difficult crowds for us to estimate because people are dispersed over several city blocks.

We reported a few weeks ago, based on an estimate from Raleigh police, that 60,000 people attended the Raleigh Christmas parade. I didn't attend this year, but I have in previous years when similar estimates were reported. And I'm skeptical.

Carter-Finley Stadium's football capacity is about 55,000. Think of the traffic and number of vehicles associated with an N.C. State football game. The Christmas parade doesn't come close to that level of activity.

It's not unusual for our crowd estimate to be different from estimates provided by others. If a crowd is small enough, a reporter might count individuals. For a larger crowd, a reporter might count one section of an auditorium and multiply by the number of sections.

Last year, reporter Thomas Goldsmith covered a rally in Wilson in support of a man who had been indicted on a murder charge. Goldsmith reported that about 450 people gathered at a church for the event.

After the story was published, a caller insisted that crowd was much larger. But Goldsmith had counted, one by one by one.