Arrow-right Camera
The Spokesman-Review Newspaper
Spokane, Washington  Est. May 19, 1883

How we came up with neighborhood numbers

Dividing a region into distinct neighborhoods – and then calculating demographics for those neighborhoods – is more art than science.

This guide is meant to be a rough overview of some of the more well-known and up-and-coming neighborhoods in and around Spokane. We’ve included demographics for each to give a snapshot, but those should be taken as approximations, not statistically rigorous analyses.

For readers who are curious about the nitty-gritty, here’s where our data came from.

Housing prices, where available, are from Zillow estimates for the median home in a given city or neighborhood. Rent prices, where available, are estimates from RentCafe.com.

Crime statistics within Spokane came from the Spokane Police Department’s CompStat numbers broken down by patrol blocks, which correspond with neighborhoods. We got numbers for the first half of 2017 and doubled them to get rough annual rates.

The U.S. Census and annual American Community Survey are the best sources of data for things like population, age and income. Census data is available for cities, ZIP codes, and smaller areas called tracts and block groups. The problem is that those areas don’t neatly line up with the neighborhood boundaries drawn by the City of Spokane.

For some neighborhoods, a single Census tract was near enough to the city-defined neighborhood boundaries that we used data for that tract. In addition to the stand-alone cities in the guide, those neighborhoods are Manito, the Perry District and Shadle.

For others, we calculated population using a GIS analysis to figure out how much of each Census block group was within the neighborhood boundaries. Those neighborhoods were Audubon/Downriver, the Garland District, Hillyard, Logan, Rockwood, Southgate and West Central.

If 45 percent of a block group was inside, say, Logan, we’d add 45 percent of that block group’s population to Logan’s total. This method assumes population is evenly distributed across a block group, which of course isn’t the case.

For median age and income, we then took the proportion of a neighborhood’s population that came from a given block group and calculated a weighted average of the Census median ages and median household incomes for those groups to get a rough median for the neighborhood.

Averaging medians is very much a statistical no-no, so those numbers should be taken as rough approximations, not true medians or exact calculations.

For a more detailed explanation with visuals, visit the Know Spokane blog at spokesman.com/blogs/know-spokane/.