Anyone who has ever met me, from my pre-kindergarten days actually, knows that I am always highly interested in technical questions and rational answers, a trait that drives some people nuts. So, while everyone is talking about the politics of Lavalas, the "184", the Democratic Convergence, etcetera, I am still curious about the basis for estimating crowds if one's intentions were not in making political points. I believe that this is a question that is important to statisticians, journalists, politicians, historians, and others.
Of course, I will choose a political example to illustrate the problem because, unfortunately, the malicious use of exaggeratedly false crowd estimates is so frequently used as a tool for campaigns of misinformation. A couple of weeks ago, I don't remember the exact date, there was, by most accounts, a huge pro-govern
ment demonstration in the streets of Port-au-Prince. I read about it from several sources. This is how I recall the reporting (I am nearly 100% sure of what I am about to write, but if I am factually wrong, please do not hesitate to correct me) :
- AlterPresse reported that thousands of people took to the street, demonstrating in the government's favor[/*:m]
- HPN reported tens of thousands of people...[/*:m]
- Associated Press and Reuters reported approximately 20,000 people...[/*:m]
- AHP reported hundreds of thousands of people...[/*:m]
thousands to hundreds of thousands, it's not just beauty that resides in the eyes of the beholder, but "truth in reporting" itself becomes quite a P.R. commodity (tell your audience either what they already want to hear or what you want to peddle to them). Was anyone concerned with telling the unvarnished truth? Hey, who the hell knows!
Given the political football going on in Haiti, I certainly do not expect things to change any time soon. It's more likely that we would discover Weapons of Mass Destruction in Iraq before this happens. (How come they have not been planted yet is beyond me... the CIA must have become afraid of its own shadow).
However, I am still intrigued... in the best of worlds, that is, one of such political stability that no news agency would feel the need to cook up numbers, what are the most accurate techniques to estimate the actual (read "as close to the truth as you can hope to get") size of a crowd in a large demonstra