One Mile Radius Residential Populations for Major Downtowns

[mhays]

I found one-mile-radius resident populations for US downtowns via CoStar property data. I guessed about locations where each downtown would have the highest number -- avoiding water/industry, including the dense fringes, but staying within the CBD.

Please suggest better addresses, and I'll update the list with any higher numbers. But stay in the CBDs.

See Canadian extrapolations at the bottom.

It says these are 2022 populations...I've emailed a CoStar rep about methodology. They might not really be 2022, or maybe they're estimates. CoStar is the top source for building data for commercial real estate people, so it could be 2020 Census plus recent completions and current occupancy rates. And (like any radius tool) how do they determine the edges, or are big chunks like block groups simply in or out?

Residents in one-mile radius, 2022:
--New York, 1008 Ave of the Americas: 205,284 (50 Central Park S was 203,678)
--San Francisco, 350 Ellis: 141,826 (605 Ellis was 150,401)
--Chicago, 222 W Erie: 134,653 (348 Lake was 112,275)
--Seattle, 906 Pine: 108,515
--Los Angeles, 601 Figueroa: 101,773
--Philadelphia, 1300 Locust: 98,852 (similar a few blocks south)
--Miami, 33 SW 2nd: 89,926 (dense area is relatively narrow)
--Boston, 165 Tremont: 84,280 (gets hit with unpopulated zones and water)
--DC, 1030 15th St NW: 80,858
--Honolulu, 550 S Beretania: 61,784
--Portland, 1331 SW Washington: 54,059
--Denver, 1570 Tremont: 53,422
--San Diego, 1130 7th: 51,218
--Minneapolis, 910 Lasalle: 46,539
--Baltimore, 250 N Calvert: 39,370
--Dallas, 1801 N Pearl: 35,348
--Austin, 920 Colorado: 33,242
--Atlanta, 384 Peachtree: 32,678 (42,176 at 675 Peachtree...if Midtown is part of CBD)
--Milwaukee, 929 N Water: 31,831
--Sacramento, 1325 15th: 30,807
--Houston, 601 Jefferson: 29,101
--Charlotte, 400 E Stonewall: 27,454
--Phoenix, 702 N Central: 25,014
--Buffalo, 707 Washington: 24,579
--Cincinnati, 44 E Court: 23,583
--New Orleans, 600 Loyola Ct.: 22,615
--Tampa, 921 N Morgan: 21,880
--Columbus, 456 E Cherry: 20,722
--Pittsburgh, 717 Grant: 19,523
--Las Vegas, 501 E 8th: 19,496
--San Antonio, 300 Convent: 17,633
--St. Louis, 1531 Pine: 17,404
--Cleveland, 1802 Chester: 17,297
--Kansas City, 601 E 12th: 15,880
--Jacksonville, 330 State: 15,052
--Detroit, 2301 Woodward: 14,325
--Memphis: 336 Monroe: 14,090
--Oklahoma City, 701 Couch: 11,181

They have Canada but only in kilometers. A two-kilo radius would be 12.56 square kilos if I'm doing it right, or 1.54x the one-mile radius.
--Toronto, 336 Yonge: 246,014 (159,749 for the average square mile)
--Vancouver, 1088 Homer: 158,022 (102,611 for the average square mile; lots of water)
--Montreal, 475 Av du President-Kennedy: 124,713 (80,982 average; dense area is fairly narrow)

[Yuri]

1 mile = 1.609 km

1sq mile = 2.59 km² (1.6 km x 1.6 km)

Got to learn those numbers by heart to post here in SSP. Someday let's hope the US will drop those unscientific units of measurement.

[mhays]

None of that's relevant here. I used radii. 2 km squared x pi for 12.56 sk or 4.85 sm.

That's vs. 3.14 sm, hence dividing by 1.54.

I could still be wrong of course.

[Skintreesnail]

1400 market would be more of the geographic center of center city Philly. Probably similar numbers though, but you would get more of the center city neighborhoods that way (might be cutting off some of fitler, old city, spring garden with your current address). Now you're probably including graduate hospital and bella Vista which are technically outside of center city (but included in greater center city).

[Easy]

Originally Posted by Yuri

1 mile = 1.609 km

1sq mile = 2.59 km² (1.6 km x 1.6 km)

Got to learn those numbers by heart to post here in SSP. Someday let's hope the US will drop those unscientific units of measurement.

You've managed to be condescending while completely misunderstanding what's happening here.

[Easy]

Interesting list. No real surprises that I see. Seattle being so close to the top is slightly surprising. I wonder where Seattle would have ranked 20 years ago? Seems like it's been booming more than most. Some sunbelts doing a little lower than I expected.

[mhays]

Originally Posted by Skintreesnail

1400 market would be more of the geographic center of center city Philly. Probably similar numbers though, but you would get more of the center city neighborhoods that way (might be cutting off some of fitler, old city, spring garden with your current address). Now you're probably including graduate hospital and bella Vista which are technically outside of center city (but included in greater center city).

Yes, but the number was lower there. I shifted it south for a higher number.

[mhays]

Originally Posted by Easy

Interesting list. No real surprises that I see. Seattle being so close to the top is slightly surprising. I wonder where Seattle would have ranked 20 years ago? Seems like it's been booming more than most. Some sunbelts doing a little lower than I expected.

Seattle benefits from a relatively-unbroken radius of residential density. I'm not surprised by its number given the total density of greater Downtown in 2020 (35k over 2.7 sm or 31k in 4.5 sm).

But I'm surprised others aren't higher than they are -- Boston, Philly, Vancouver... Boston and Vancouver are heavily influenced by water and parks.

[Skintreesnail]

Yeah, south Philly definitely packs a punch density wise so i guess I'm not surprised. There's a lot of development going on in callow hill, so maybe it'll change in a few years.

[lio45]

Originally Posted by mhays

None of that's relevant here. I used radii. 2 km squared x pi for 12.56 sk or 4.85 sm.

That's vs. 3.14 sm, hence dividing by 1.54.

I could still be wrong of course.

And you are. A one-mile-radius circle has an area of 8.13 square kilometers.

Edit: Oh, I see that you’re stuck using kilometers because that’s how the data is; then yes, you’ll want to only take the inner 1/1.54 fraction to try to approach apples-to-apples.

Not perfect, as it assumes uniform density, but likely not too far off.

[mhays]

That matches what I used. Divide 12.56 by 1.54 and you get a little over 8.15 (rounded).

[Yuri]

Originally Posted by mhays

None of that's relevant here. I used radii. 2 km squared x pi for 12.56 sk or 4.85 sm.

That's vs. 3.14 sm, hence dividing by 1.54.

I could still be wrong of course.

I see. You won't the same area for Canada and the US.

Originally Posted by Easy

You've managed to be condescending while completely misunderstanding what's happening here.

Condescending?! The colleague said he was not sure about the number and I posted a formula. That's it. We use this forum to exchange info. There's no shame on asking and answering stuff. Everybody does both all the time.

[lio45]

Originally Posted by Easy

You've managed to be condescending while completely misunderstanding what's happening here.

Same here, but then I paid attention and understood.

[Easy]

Originally Posted by Yuri

Condescending?! The colleague said he was not sure about the number and I posted a formula. That's it.

Yes. You made a condescending mark about "unscienctific" while evidently still not grasping that the formulas for calculating the areas of a circle and a square are different.

[Yuri]

Originally Posted by Easy

Yes. You made a condescending mark about "science" while evidently still not grasping that the calculations for the areas of a circle and a square are different.

I was joking about the US imperial system, which is always subjected to similar jokes. Americans also make fun of the metric system. It was not directed to the colleague.

Areas of squares and circles are calculated differently, that's obvious. However, 1 sq mile (be it in a square or in a circle shape) will always be 2.59 km².

[lio45]

Originally Posted by mhays

That matches what I used. Divide 12.56 by 1.54 and you get a little over 8.15 (rounded).

In case you didn’t think of it, to test the quality of your approximation, you could compare it with the other way of getting a mile radius from the Canadian data: take the 1 km radius, and multiply by 2.59.

[badrunner]

Originally Posted by lio45

In case you didn’t think of it, to test the quality of your approximation, you could compare it with the other way of getting a mile radius from the Canadian data: take the 1 km radius, and multiply by 2.59.

That would be like checking the accuracy of a measurement with a less accurate measurement. 1.54 < 2.59.

[lio45]

Originally Posted by badrunner

That would be like checking the accuracy of a measurement with a less accurate measurement. 1.54 < 2.59.

In science, checking the validity of an assumption by comparing it to an independent process that should, if the assumption is valid, yield the same result, is a perfectly logical and very common method.

If they are that different, then it tells us that the numbers mhays is seeking are almost certainly in reality a hybrid of the two values - which is perfectly attainable, for the record. Just take the average, that’s already an improvement - or if you insist, weigh the value extrapolated from the two km circle more than the one from the one km one (weighted by closeness to the one mile circle; again easy).

[Easy]

Originally Posted by lio45

In science, checking the validity of an assumption by comparing it to an independent process that should, if the assumption is valid, yield the same result, is a perfectly logical and very common method.

If they are that different, then it tells us that the numbers mhays is seeking are almost certainly in reality a hybrid of the two values - which is perfectly attainable, for the record. Just take the average, that’s already an improvement - or if you insist, weigh the value extrapolated from the two km circle more than the one from the one km one (weighted by closeness to the one mile circle; again easy).

My guess is that the program only accepts integers or mhays word have converted 1 mile to kilometers from the beginning.

[mhays]

I used this site to figure out the area of a 2-km radius.

CoStar isn't a demographics site per se. It's just trying to show marketability for each building.

They only have locally-appropriate radius options, so only miles in the US and only kms in Canada. I wish they provided both for both.