Tom Meagher bio photo

Tom Meagher

News nerd.

Twitter LinkedIn Github

How to get a head count at the parade

This is another in the series of “Best Practices” posts from my archives from the Herald News. This was also written by Jonathan Maslow, and offers some solid street reporting tips for counting the crowd at any large public event.

News reporting often involves telling how many people took part in or attended an event. Sometimes the seating capacity is known, as in the case of a sports stadium. Sometimes an official estimates the crowd size, sometimes the organizers estimate. Sometimes they don’t agree. Sometimes no one estimates.

It’s a good practice for the reporter to do his or her own estimation in all cases.

For inside events, a quick head count with a qualifying “about” is good. (“about 75 people attended the planning board session.”) The tried and true method of estimating crowd size is as follows: 1) make a careful measurement of the area in which the event takes place. Simply put, that’s length times width equals square feet. In some cases, it may take a slightly more sophisticated algorithm (the crowd covers the sidewalks on six blocks. Ask officials how long each block is, walk off the width of sidewalk. Multiply).

2) Decide whether it’s a dense crowd or a loose crowd. People packed into a subway car, for example, are a dense crowd and occupy 2 square feet per person. A loose crowd, such as a parade of journalists on May Day, has one person per five square feet.

3) Divide the area by the crowd density, and you’ve got your estimated crowd. Example: A political rally takes place in the square outside City Hall. You pace it off and find it’s 150 feet long and 100 feet wide, or 15,000 square feet. You wade into the crowd to test the density. Some places toward the front are packed, but in the back it’s loose. You choose a density in between of one person per three square feet. Divide 15,000 by 3 = 5,000 people.

To recap:

area: 150 x 100 = 15,000 estimate of density: 3 sq. feet/per person 15,000 divided by 3 = 5,000 attendees

If AP is covering an event, check your estimate against theirs, which is based on aerial/satellite photography and should be accurate. If yours and theirs are at great odds, dig in your heels and insist the wire services are an ass.

Note: [Then] Passaic County Sherriff’s communications man Bill Maer says the county uses a slight variation:

1) Count how many people are actually in a 10 x 10 foot area (that’s 100 square feet).

2) Measure on foot or get a good estimate of the size of the entire area.

3) divide the total area by 100 (the number of 10 by 10 squares)

4) Multiply the result by how many people are in the sample 10 x 10 foot square.

Personally, I think this adds an unnecessary step, but it does make the reckoning of density a bit more accurate and hence, perhaps, the crowd estimate itself.

  • Jonathan Maslow, Herald News, Nov. 2003