[OldNorth] Population density reported on Census website accurate to 10% (for Block)

[Z Smith]

Everybody,

John and I have spoken by phone about the finer points of calculating 
population density.  Here's the scoop:

The Census data he was able to download 2 years ago gave Block area in 
square miles to just two decimal places, and since a typical Block in Old 
North is just 0.00665 or 0.00550 square miles, the old datasets would show 
these all as just 0.01 square miles, and dividing the population of a block 
into 0.01 square miles would make the population density numbers off by 
something like 40%.

The current website lets you view & download the population density and the 
population for each block.  Dividing one by the other gives you the area 
they assumed.   I did this and found that the areas they are using for 
blocks appear to be stored to 3 decimal places-- that is, they use 0.007 or 
0.006 or 0.005 square miles as the area of various blocks.   This introduces 
an error of up to 10% in the calculated density.

For my work, I have in fact measured the dimensions of the blocks and 
streets of Old North and checked them against the city maps.   From this 
work I was able to calculate the "true" density and compare it with the 
values reported by the Census website--again, they match to within 10%. I 
will be happy to send the spreadseet with all this data to anyone who'd 
actually like to see it.

So, the bottom line is that the densities you find clicking around on the 
Census maps are probably good to + or - 10% for individual Blocks.  The 
relative error is less when you look at groups of Blocks or whole Census 
Tracts.

Z

_________________________________________________________________
Enjoy a special introductory offer for dial-up Internet access — limited 
time only! http://join.msn.com/?page=dept/dialup