Check out this article about First Solar purchasing a piece of property.
First Solar Inc. of Tempe is buying about 635 acres of land owned by Mesa in Pinal County, where the company plans to build an electrical-generating station for Salt River Project.
The purchase price is about $9.8 million, or $15,433 per acre.
What's wrong with the estimate? The company is buying "about" 635 acres for "$9.8 million." So how accurate can the final number be? Not very. But notice the conclusion...the price is $15,433 per acre. This is obviously wrong because we don't have enough information on the two factors--size and price--to do any more than make a rough calculation of price per acre. The best the author can do is say the price is "about $15,000 an acre."
Why does it matter? Am I just being picky? It matters because it's a great example of the type of false confidence that policy analysts, politicians and journalists often display.
Policy makers like to take inputs that are difficult to estimate--popularity of various sports venues, convention center demand, or macro issues like "stimulus" spending and Quantitative Easing-- run them through various "multipliers" and "levelers" then assume some sort of discount rate and come up with the present value of a 20 year stream of revenue projections. Oftentimes the final "answer" is very specific, yet the inputs are highly speculative. The media then dutifully report these outputs as fact. That's why it's instructive that the Republic's Gary Nelson was willing to say that the cost of the First Solar purchase was $15,433 an acre.
Here's a great example of the futility of this type of analysis. Take a look at this 2003 article about the City of Glendale's great idea for a new sports arena...
City officials estimate the arena will more than pay for itself and even greater profits will be realized with the Westgate development. The Coyotes will pay at least $20 million rent during their 30-year lease and sales taxes from tickets, concessions and suites are predicted to bring in $43 million. The big money generator is parking, which officials figure will generate more than $130 million in revenue for the city over the life of the Coyotes' contract.
Golly, I bet that $43 million number was comforting--take the numer of hotdogs, pretzles, cokes and beer per game for the next 30 years, factor in inflation and the projected sales tax rate then take the present value and you get a nice number. Of, course it's actually a nice meaningless number, but it's so comforting because it's so precise.
Now Westgate is in foreclosure, the team is ready to leave, Glendale is struggling to pay its debt and the $43 million has been shown to be an illusion.
It's too bad that no one in the media looked at the figure skeptically.
So next time that you read that a big sports event is going to bring $212 million into the Valley, or a new program will create 150,000 jobs. Take a moment and remind yourself that there's one thing you can definately say about the number. It's wrong.