My wife and I were particularly enthusiastic about the Clifton Corridor proposal, one of the bigger projects that would have been funded by the tax, and voted yes. A whopping 63 percent of voters across the region didn’t quite share our enthusiasm, however, and the measure went down in flames despite a multi-million dollar campaign by supporters.

It should be noted that both the funding mechanism itself and the list of proposed projects were deeply flawed, and that the whole thing was almost certainly doomed to fail from the very moment it was conceived. Consider that the referendum failed 51 to 49 percent even in Fulton and DeKalb counties, the only two metro counties in which residents have for more than 40 years paid an extra 1 percent sales tax to support the MARTA transit system.

Consider, too, that in 2008 President Obama won 79 percent of the vote in DeKalb County and 67 percent in Fulton. Then consider that among the groups opposing T-SPLOST were the Sierra Club and Georgia NAACP. Now consider how stupid this AJC story was for crediting the Tea Party with a “big win.“

Anyway, I live in DeKalb County and was curious how the vote broke geographically, so I put together the interactive map of precinct-level results below. Green precincts are where the majority of voters backed T-SPLOST, while in red areas the majority voted no. The darker the color, the more pronounced the majority.

Mash your mouse right here to see a larger version of the map.

Support for T-SPLOST was highest in the Epworth precinct in the city of Atlanta, at 79 percent, and lowest in the Kingsley Elementary precinct in Dunwoody, where 71 percent votied against the measure. In the city of Decatur, 41 percent of voters cast a T-SPLOST ballot, with 72 percent voting yes. Turnout was 32 percent in the DeKalb County portion of the city of Atlanta, with 69 percent voting yes.

And here’s a bonus map shaded by T-SPLOST vote turnout. Darker precincts had higher turnout, lighter precincts lower. The countywide rate was 31 percent, with a high of 79 percent in the Medlock Elementary precinct and a low of 5 percent in the Panola Road and Emory Road precincts.

Mash your mouse right here to see a larger version of the map.

Let me know if you see any mistakes in the maps or data, or if you’re interested getting your hands on the precinct boundary maps or precinct results I used here.

What about the other nine counties in the Atlanta T-SPLOST region? Well, I have precinct results for Fulton County, but haven’t even looked at them yet. I also haven’t even bothered tracking down results from the other eight counties. So if you want to see maps and info like this for the rest of the metro area, let me know. Better yet, send me the data, or at least a link to it or contact info for its custodian. Better still, send me some money.

I’m serious about that last part.

Don’t forget to send me some money.

UPDATE, Aug. 17: Oops. Just noticed a note on the DeKalb elections site about combined precincts. The Emory Road and Druid Hills High precincts were combined at one location, so the Emory Road results are included in the Druid Hills High results. Same goes for Panola Road and Miller Grove High, with the former being included in the latter’s results. And Scott precinct results were included in the Medlock Elementary results.

That explains both the unusually high and unusually low turnout figures for the precincts above. The vote totals you do see for the combined precincts were mainly in the form of absentee ballots. I might try to combine the actual precinct boundaries and update the maps, but I can’t make any promises. Not until you send me some money, anyway.

Curious about MuckRock, and nursing my own long-standing grudge against another piece of contractor-built government software, in January I requested “the total cost of and the name of the company that developed the current version of the American FactFinder web application.”

Then, to demonstrate my unwavering commitment to relentlessly pursuing this outrageous squandering of taxpayer money, I instantly forgot about the whole thing. In the meantime, just three weeks after my request, the Census Bureau responded with the exact information I was after:

The name of the company that developed the current version of the American FactFinder web application is IBM U.S. Federal and the total $33,340,681.00.

Did you catch the part where it said the Census Bureau paid IBM $33.3 million for the poo bucket that is the American FactFinder? Ha ha. That was my favorite part, too.

As I said, I forgot about the request until today, signed into MuckRock and found the Census Bureau’s response there waiting for me since February. So kudos to the Census Bureau for a refreshingly prompt response, and to MuckRock for handling the FOIA process so smoothly on behalf of an absentee requester.

Now, about that $33.3 million. As someone who worked extensively with data from the Census Bureau during my 11-year career as a journalist, and who continues to tinker with it today purely for my own amusement, $33.3 million seems like an APPALLINGLY LARGE AMOUNT OF MONEY TO WASTE ON THE USELESS BOONDOGGLE THAT IS THE AMERICAN FACTFINDER, WHICH, AS ANYONE WHO HAS TRIED TO USE THE THING WILL NO DOUBT ATTEST, NO DOUBT WHILE SHOUTING OR TYPING IN ALL CAPS, NO DOUBT WHILE WAVING THEIR ARMS ABOUT IN A MANNER THAT COULD BE DESCRIBED AS COMICAL IF NOT FOR THE CRAZED LOOK IN THEIR EYES, IS A USELESS BOONDOGGLE THAT IS USELESS AND A BOONDOGGLE.

Or maybe I’m wrong and that’s a bargain price! For this:

Here’s the response the Census Bureau sent MuckRock:

Update, July 12, 2012: Here’s a response from the official Census Bureau blog, “FactFinder is More than Just a Website.” It includes an admission that “the American Customer Satisfaction Index scores for FactFinder results mirror the comments from our friends in the blogosphere, with our data getting higher rankings than functionality and navigation. But as we make improvements, those functionality scores are rising dramatically.”

We’re friends!

Also worth highlighting from the post: “But if you think you can do better, you’ll soon have the opportunity. The imminent release of our API will let developers build apps tailored to meet the needs of any and all audiences.”

So I just made some maps of the 28-county Atlanta metro area using block-level data from the 2010 Census instead.

The first map shows the black population. The shading represents, from light blue on the low end to dark blue on the high end, the percentage of each block’s population that is non-Hispanic black. The height of each block shows the number of people in each block who are non-Hispanic black.

(Click the map for the full-size image.)

Same deal, except this one shows the Hispanic population. Again, shading represents share of the population, height shows the count.

(Click the map for the full-size image.)

And one more time for the Asian population.

(Click the map for the full-size image.)

A similar map showing the non-Hispanic white population doesn’t work all that well visually, so I left it out. Just sort of imagine the inverse of the first map above. Or, better yet, have a look at this block-level 3D map of the non-Hispanic white population in the Atlanta metro area I did awhile ago using 2000 Census data (it’s a PDF, by the way).

Hey, in fact, here are a bunch of other maps along the same lines that you may (seriously, you have my permission) enjoy. They’re all PDFs, some of the multiple-megabyte variety, so click accordingly.

• 2D map of the majority race by block in the Atlanta metro area in 2000
• A slightly brighter but otherwise mostly identical version of the previous map
• 2D map of the majority race by block for all of Florida in 2000, a particularly big file
• 3D map of the black population by block in the Atlanta metro area in 2000
• 3D map of the Hispanic population by block in the Atlanta metro area in 2000
• 3D map of the black population in Palm Beach County, Florida, with 2000 data adjusted to 2010 tabulation blocks
• 3D map of the black population in Palm Beach County, Florida, with 2010 data in 2010 blocks
• 3D map of the Hispanic population of Palm Beach County, Florida, with 2000 data adjusted to 2010 tabulation blocks
• 3D map of the Hispanic population of Palm Beach County, Florida with 2010 data in 2010 blocks
• 3D map of the block-level change in total population from 2000 to 2010 in Palm Beach County, Florida

Wasn’t that fun? Now, as promised, I have one last treat for you, an interactive block-level bonus fun time party map of DeKalb County. The shading reflects the percentage of each block’s population that is non-Hispanic black, from dark purple for a very small share of the population to dark green for a very large share. Blocks shaded red have no residents.

The map is centered on Decatur, but zoom out or pan about for all of DeKalb County. Click on any block to see its 2010 population and basic racial and ethnic breakdown.

Want to see a much larger version of the interactive map? Just mash your mouse down here and we’ll go there in a new window or tab together, hand-in-hand, skipping all the way.

Are we done yet? The heck we are. If you liked everything above enough that you kept reading to this point, I bet you would like to have a look at these county-level 3D maps of 2004 and 2008 presidential election results.

And if you like those, I should probably just come over to your house so we can hang out, talk about cats, and compare the features of our smartphones.

For no particular reason apart from the fact that this is the sort of thing I do for fun, I have prepared the following charts that show monthly price and sales trends in Palm Beach County’s residential real estate market from January 1995 to June 2010. (If you’re interested in similar information but at the neighborhood level, you can see pretty much the same information for more than 3,000 Palm Beach County neighborhoods over at my day job.)

Careful readers might recall that I did something similarly purposeless in 2008, when I made a series of 3D maps showing county-level results from the 2008 and 2004 presidential elections. Careless readers inexplicably put their laptop in the refrigerator last month, and haven’t been online since. Extremely tall and athletically gifted readers blessed with youthful energy and good health are currently preparing for the forthcoming professional basketball season.

To see the full-size version of the charts below, simply click on the image. To avoid seeing the full-size charts, simply do nothing and continue looking (with your eyes) at the small versions currently before you. You’ll probably want to then contact me via fax or letter, and I will arrive at your house in no less than seven (10) business days to further assist you.

The first chart shows the monthly median sale price by housing type. The single-family home median sale price peaked in Palm Beach County in December 2005 at $418,310. By December 2009 it was down to $249,000, and continued declining through the first six months of 2010.

(click for full-size image)

And what about this handsome devil? Oh, it just shows the total number of qualified or arms-length residential transactions by price group. And looks good doing it.

(click for full-size image)

Similar to the one above, but this one shows each price group’s share of residential sales.

(click for full-size image)

Behold, the year-over-year change in the total number of residential transactions. Got that? January 1996 compared to January 1995, February 1996 compared to February 1995, etc. Listen, this isn’t rocket science. Looks like some good news the past 18 months or so, right? Sales finally picking up again?

(click for full-size image)

Similar to the one above, but this one shows the year-over-year change in median sale price for all housing types. So, maybe not such good news, then. Fact alert: Median prices increased 90 consecutive months from April 1999 to September 2006, then decreased 45 consecutive months from October 2006 though the end of the period covered by my charts, June 2010.

(click for full-size image)

So there you have it. Want the raw data on which these charts are based? Or the even rawer data on which the data on which these data these charts are based are based? TOUGH LUCK HIPPIE, with your dangerous Web 3.0 ideas and your Arctic Monkeys and fancy PlayStation 2 consoles. But really, if you want it, get in touch and I’ll think about it, then maybe we can go to the movies together a few times to get to know each other a little better, then maybe I’ll send it to you.

Here’s how Barack Obama did. The four color ranges represent, from light to dark and in simple quartiles, the percentage of total votes Obama received in each county. Height shows the number of votes Obama received.

(Click the map for the full-size image.)

Here’s how John McCain did. Again, four color ranges representing share of votes, and height showing the number of McCain votes.

(Click the map for the full-size image.)

Compare those maps to their equivalents from the 2004 presidential election. John Kerry in blue, George W. Bush in red.

(Click the map for the full-size image.)

(Click the map for the full-size image.)

The next two are slightly more complicated. I’m not convinced they actually work terribly well, but I made them, so here they are. The first shows the difference between Obama’s 2008 performance and Kerry’s in 2004. There are six color ranges that represent the percentage difference between the number of votes received by Obama and Kerry. Three red shades represent declines from 2004 to 2008. From dark to light red: -65% to -40%, -39% to -20% and -19% to 0. Three green shades represent increases. From light to dark green: 0 to +20%, +21% to +40% and +41% to +110%. Each county’s height shows the simple numerical difference in the number of Democratic votes in 2004 and 2008. A bit hard to tell in the small version of the map below, but negative values actually do dip below the, um, earth’s surface.

(Click the map for the full-size image.)

And here’s the same deal comparing McCain’s 2008 performance to Bush’s in 2004. From dark to light red: -67% to -40%, -39% to -20% and -19% to 0. And from light to dark green: 0 to +20%, +21% to +40% and +41% to +56%. Anything catch your eye when you compare this map to the one above? You might have to look carefully, as it’s very subtle.

(Click the map for the full-size image.)

Last set. As in the first set, these are again shaded by the percentage of votes received by Obama and McCain. Four colors, simple quartiles, light is low, dark is high. But in this case height represents the difference between Obama’s total votes and McCain’s. Obviously, these maps mirror each other, but I like looking at them separately.

(Click the map for the full-size image.)

And from McCain’s perspective.

(Click the map for the full-size image.)

After the jump are the details for Microsoft visitors between February 13, 2007, which I consider this site’s date of birth, and October 16, 2008, which I consider, somewhat controversially, to be yesterday. The numbers don’t add up to 100 percent because they just don’t, and that’s just something we’re all going to have to live with, as there really isn’t anything we can do about it.

MICROSOFTIE BROWSERS

• Internet Explorer: 86.5%
• Firefox: 7.3%
• Safari: 3.1%
• Chrome: 1.0%

MICROSOFTIE OPERATING SYSTEMS

• Windows: 93.8%
• Mac: 4.2%

What about visitors from Google? You’re not going to believe your luck, because I have those numbers right here.

GOOGLER BROWSERS

• Firefox: 48.7%
• Internet Explorer: 25.6%
• Mozilla: 10.3%
• Safari: 5.1%
• Opera: 2.6%

GOOGLER OPERATIONS SYTEMS

• Windows: 76.9%
• Mac: 10.3%
• Linux: 5.1%

Matt started Stormpulse before joining the Backyard Post team at The Palm Beach Post, so it’s cool to see the two finally come together. And though I have absolutely nothing to do with Stormpulse, seeing its mapping software embedded on The Post is personally gratifying because of the crazy, fortuitous way that I met Matt.

I was still talking ears off and hustlin’ Backyard Post around the office on April 26, 2007 when I had my first conversation with The Post’s online director about hiring a professional web developer. Literally seconds after I got back to my desk following that meeting I ran across this post about Stormpulse on the predecessor to Jason Spalding’s current site, Find GIS. That pointed me to this article at GISuser.com, and finally on to the site itself. It blew me away, so I started digging around to find out more about the people who made it. I noticed one of the guys lived in West Palm Beach, left a comment on the blog asking him to get in touch and casually mentioned later that I had just been in a meeting about hiring a programmer. Two months passed, Matt told me he was job-hunting, I went completely mental and flipped out on just about everyone in the building because he had other offers almost instantaneously, and either because of or in spite of my crazed urgency, we ended up hiring the guy. A few months later we also hired Matt’s high school classmate Peter Sheats, and here we are.

And about that “Coming soon to a site near you?” bit from the headline. The even cooler Stormpulse news is that anyone interested in embedding their tracking map on your own site can now do so. Check out the Stormpulse blog for details.

Then there’s my own neighborhood. Here’s the entrance I use, as seen in Street View:

View Larger Map

Buildings and stuff, right? Follow the jump to see how my neighborhood looks in Google Map’s outdated aerial view. The vantage point above will be in the lower right corner of this map, facing northwest.

View Larger Map

Not cool, right? Anyway, here are a couple more Palm Beach County treats. The west-facing view of The Palm Beach Post’s main office, as also featured in this previous post.

View Larger Map

And the private driveway leading to Rush Limbaugh’s oceanfront mansion on Palm Beach. Not a whole lot to see, though you can peek over the hedges in Virtual Earth.

View Larger Map

Seems like about half of the comments I’ve left on WordPress-powered blogs running Akismet lately have been diverted into the spam queue. Maybe it knows something I don’t, but I’ll just put out a friendly reminder to check your spam folder every now and then for legitimate comments.