From the Memphis Business Journal of October 10, 2012, "Fastest Growing Cities":
Each of the top five fastest growing cities in the Memphis area is in DeSoto County according to population growth percentages from 2010 to 2011. According to the Census Bureau, each of those cities — click here to see the top five — grew its population by more than 1 percent in a year’s time.
How do you reconstruct these results?
The American Community Survey comes out annually with an estimate of population size based on survey results. Because the ACS uses sampling, instead of a complete population count like the every 10 year census, it also reports on how good the estimates are expected to be.
Let's look for "fastest growing cities in Michigan, 2010 to 2011", and also for "fastest shrinking cities in Michigan, 2010 to 2011", where you're looking for population loss.
The tables I found were from the American Factfinder census database, where I searched for "Population" and go back table B01003, "Population". Two deep links will take you to the queries I ran, one for 2011 and the other for 2010. I limited the geography to "places within Michigan", so that I'd get cities. (And that will come back to be a problem - read on...) Because 1-year ACS estimates only report on areas of a certain size, I only got back data from a limited number of cities, 16 in all. I then pasted the results into a Google Spreadsheet, and in that spreadsheet computed the difference between 2010 and 2011 both as an absolute value and as a percentage.
What came back was this:
Detroit lost the most population, with an estimate of 5,270 or 0.75% population loss.
Grand Rapids gained the most population, with est. gain of 1,599 or 0.84% gain.
Dearborn had the greatest percentage loss, losing 764 or 0.79% loss.
Wyoming had the greatest percentage gain, gaining 630 or 0.86% gain.
Both Detroit and Dearborn are in Wayne County, and both Grand Rapids and Wyoming are in Kent County; that suggests that another interesting point of comparison is population loss or gain by county. Continuing...
2011 by county data here.
2010 by county data here.
Run the same analysis, comparing current versus previous year on a county basis. We get more area in the sample, because the one year estimates go down to about 70,000 people minimum per each.
Fastest growing is Grand Traverse County, gaining 1306 people or 1.48%.
Biggest growth is Oakland County, gaining 7133 people or 0.59%.
Fastest shrinking is Shiawassee County, losing 765 people or 1.10%.
Largest loss is Wayne County, losing 13,768 people or 0.76%.
Kent County, for comparison, had two of the fastest growing cities, and gained 5234 people or 0.86%.
--
Now, did I get this right?
I used ACS data for both of them, rather than using 2010 decennial census data. I think that's a reasonable point of comparison to choose the same survey instrument.
The geography I picked was the "place" and the "county" level. In Michigan there's also a "county subdivision" level, and if you do that, you also get "charter townships" - and some of these have sizable populations. So, I didn't get it exactly right, and to do it properly all over again I'd run numbers for Canton Township and Shelby Township and a few others. The 2011 data is here and the 2010 data is here. So this analysis stays a draft! until I can fix that error....and the results:
None of the townships were the biggest loser or biggest winner, so the overall analysis is still OK.
Several townships saw losses by percentage that were notable (over 0.75%): Canton was projected to lose almost as much by percentage as Detroit.
This gives me pause, though - when I look at the numbers, all of the cities in a county have about the same level of gain or loss, expressed as a percentage. All of the Macomb County population changes are within a margin of error, and they're all tightly clustered around the county average. Is this telling me that the one-year ACS estimates are really only accurate at the county level, and can't be used to distinguish among gain or loss rates within a county? That would be an interesting and useful result.
Time to query further, I think, but not tonight. I need to reconsider whether I should use 2010 census data instead of 2010 ACS data for this, because it might be more accurate.
--
As it turns out, the way to this answer should be a lot simpler than digging through individual detail results, and it is.
The Census has a Population Estimates page, which publishes in a simple, compact, tabular format the same data I'm trying to extract from the detail pages of American Factfinder. Hooray!
See
- Cumulative Estimates of Resident Population Change for Incorporated Places over 50,000, Ranked by Percent Change: April 1, 2010 to July 1, 2011 (SUB-EST2011-02) [XLS - 54k] | [CSV - 17k]
for the precise data. This has a very compact CSV file that is already sorted neatly by percent population change, using a standard time span. Looking just for Michigan in that file, you get this complete result:
412,Wyoming city,Michigan,"72,125","72,833",708,1.0, 418,Grand Rapids city,Michigan,"188,041","189,815","1,774",0.9, 438,Ann Arbor city,Michigan,"113,934","114,925",991,0.9, 525,Troy city,Michigan,"80,980","81,508",528,0.7, 527,Novi city,Michigan,"55,224","55,583",359,0.7, 528,Farmington Hills city,Michigan,"79,740","80,258",518,0.6, 530,Royal Oak city,Michigan,"57,236","57,607",371,0.6, 531,Kalamazoo city,Michigan,"74,262","74,743",481,0.6, 533,Southfield city,Michigan,"71,739","72,201",462,0.6, 534,Rochester Hills city,Michigan,"70,995","71,452",457,0.6, 537,Pontiac city,Michigan,"59,515","59,887",372,0.6, 612,Lansing city,Michigan,"114,297","114,605",308,0.3, 629,Sterling Heights city,Michigan,"129,699","129,880",181,0.1, 630,Warren city,Michigan,"134,056","134,243",187,0.1, 631,St. Clair Shores city,Michigan,"59,715","59,798",83,0.1, 700,Battle Creek city,Michigan,"52,347","52,093",-254,-0.5, 701,Saginaw city,Michigan,"51,508","51,230",-278,-0.5, 709,Flint city,Michigan,"102,434","101,558",-876,-0.9, 710,Detroit city,Michigan,"713,777","706,585","-7,192",-1.0, 711,Livonia city,Michigan,"96,942","95,958",-984,-1.0, 712,Westland city,Michigan,"84,094","83,239",-855,-1.0, 713,Taylor city,Michigan,"63,131","62,489",-642,-1.0, 714,Dearborn Heights city,Michigan,"57,774","57,186",-588,-1.0, 715,Dearborn city,Michigan,"98,153","97,144","-1,009",-1.0,
If you know a little geography, you'll recognize that all six of the fastest losing cities in Michigan are in Wayne County; if you know a little more about geography, it's that this list doesn't show Michigan charter townships! Still, it's the same relative rankings, notably at a reported precision of 0.1% rather than the 0.01% I gave earlier.
After some detour, I think we have our answer.
Comments