"Media markets?" you might be sputtering. "But I just got done redistricting the House into 1,000 seats using the Wyoming rule, calculated their PVIs, and then gave them all ridings names!" Yes, here's one more way of breaking down the country into component parts, a favorite activity at Daily Kos Elections if the recent diary list is any indication.
This project, of course, has no immediate bearing on our electoral prospects. Neverthless, I feel media markets are a really interesting level of analysis that fill in a gap between states and congressional districts (or counties), and I think they're worth taking seriously. I've been compiling data on media markets now so that we have the tools available to better discuss the money behind ad-buying decisions later, as the campaign season starts to really go full tilt. (Those of you who were around for the Swing State Project days may remember the bang for the buck index, which looked at the media-costs-related efficiencies of investing in different netroots candidates, and which is something I definitely want to reprise this year... in the meantime, I thought I'd share some of the demographic highlights.)
On the one hand, media markets are a more granular sort than states (there are 210 of them), without getting too granular, as with counties (of which there are more than 3,000), or even CDs, which at 435 can get a little mind-numbing. (And unlike CDs, they're compact, rather than, usually, gerrymandered to maximize political advantage.) And on the other hand, they're a lot more culturally cohesive than states, differentiating parts of large states that encompass a diverse array of demographics... and they also reflect the fact that those cultural/demographic affinity clusters don't usually honor state lines, especially where major metropolitan areas sit on both sides of large rivers.
If you've ever tried to research this topic, you probably know how hard it is to find any definite information even about how many people live in a market, let alone what its rates per point are. Luckily, with a little searching you can find federal government data on what county goes in what market, and from there you can spreadsheet county data to get demographic and/or political data per market. (That link contains a map of each market; here's one less cumbersome map that puts it all in one place, though it seems to have some minor errors compared with the FCC data.)
Why is knowing that stuff important? Because the population of a media market indicates how expensive ad time will be in that particular market; in other words, you pay more to reach more eyeballs. At the same time, though, the market demographics also matter, because advertisers will pay more per eyeball to reach a market where people are likelier to buy the subject of the advertisement. Obviously not every small-dollar contributor can become an expert media buyer, but it can help us make better investment decisions (where choosing which netroots candidates get our small dollars is, essentially, an investment).
So, let's look at the political numbers first; over the fold, we'll talk about demographics a little too. Let's start with the bluest markets:
Market |
2008 vote |
2004 vote |
PVI |
San Francisco, CA |
73/24 |
69/29 |
D+21 |
Monterey, CA |
72/26 |
66/33 |
D+19 |
Springfield, MA |
65/32 |
64/35 |
D+15 |
Madison, WI |
68/31 |
60/39 |
D+13 |
Laredo, TX |
71/28 |
57/43 |
D+13 |
Chicago, IL |
67/32 |
60/40 |
D+13 |
Honolulu, HI |
72/27 |
54/45 |
D+12 |
Duluth, MN |
63/35 |
62/37 |
D+12 |
Providence, RI |
62/36 |
61/38 |
D+11 |
Washington, DC |
65/34 |
58/41 |
D+11 |
And here are the reddest markets. You're probably noticing that most of them are in rural Texas. Some of you might also notice that there's a huge population disparity between the 10 bluest markets (which have a total population of 29 million) and the 10 reddest (total population 4 million... if you're wondering where you find the reddest market with a population of more than 1 million by itself, that'd be Salt Lake City at R+21, at #13).
Market |
2008 vote |
2004 vote |
PVI |
Amarillo, TX |
22/77 |
20/80 |
R+30 |
Abilene, TX |
24/75 |
22/77 |
R+28 |
Odessa, TX |
26/73 |
23/77 |
R+27 |
San Angelo, TX |
27/72 |
23/76 |
R+26 |
Lubbock, TX |
29/70 |
23/77 |
R+25 |
Dothan, AL |
28/71 |
26/74 |
R+24 |
Idaho Falls, ID |
31/66 |
26/73 |
R+22 |
Wichita Falls, TX |
30/70 |
29/71 |
R+22 |
Tyler, TX |
30/70 |
29/70 |
R+22 |
Panama City, FL |
29/70 |
30/69 |
R+22 |
Here are the swingiest markets, which I'm defining as the ones with PVIs closest to even. (You may be asking "Milwaukee? Really?" Sure, Milwaukee County is the biggest clump of Democratic voters in Wisconsin. But the market also includes all of the Circle of Ignorance: Waukesha, Washington, and Ozaukee Counties, whose hard-core Republicanism balances out Milwaukee proper. As you can see in the top 10 bluest list above, it's Madison that pushes Wisconsin over the top in the blue direction.)
Market |
2008 vote |
2004 vote |
PVI |
San Diego, CA |
54/44 |
48/50 |
D+1 |
Wausau, WI |
55/43 |
48/51 |
D+1 |
Rochester, NY |
55/44 |
48/51 |
D+1 |
Tucson, AZ |
51/47 |
51/48 |
D+1 |
Tallahassee, FL |
51/48 |
51/48 |
D+1 |
Des Moines, IA |
52/46 |
48/51 |
R+0 |
Milwaukee, WI |
53/46 |
48/51 |
R+0 |
Norfolk, VA |
55/44 |
46/53 |
R+0 |
Austin, TX |
53/45 |
47/52 |
R+0 |
Marquette, MI |
52/46 |
48/51 |
R+1 |
More over the flip...
A few more electoral numbers; let's look at the 2004-to-2008 trend and see which markets moved most toward us and away from us. I'm defining "trend" as the difference between the 2004 Kerry percentage and the 2008 Obama percentage. The numbers won't be surprising to those people who followed which states moved the most toward the blue column:
Market |
2008 vote |
2004 vote |
Trend |
Honolulu, HI |
72/27 |
54/45 |
+18 |
Lafayette, IN |
54/44 |
39/60 |
+15 |
Laredo, TX |
71/28 |
57/43 |
+14 |
Harlingen, TX |
68/31 |
54/46 |
+14 |
Indianapolis, IN |
50/49 |
38/62 |
+12 |
Ft. Wayne, IN |
43/55 |
32/67 |
+11 |
Omaha, NE |
46/52 |
36/63 |
+10 |
South Bend, IN |
49/49 |
39/60 |
+10 |
Fargo, ND |
50/48 |
40/58 |
+10 |
Terre Haute, IN |
47/51 |
38/62 |
+9 |
Likewise, the markets that trended away from the Dems the fastest should look familiar:
Market |
2008 vote |
2004 vote |
Trend |
Jonesboro, AR |
36/60 |
48/51 |
-11 |
Little Rock, AR |
41/57 |
47/52 |
-6 |
Lake Charles, LA |
34/64 |
39/60 |
-5 |
New Orleans, LA |
43/55 |
48/51 |
-5 |
Sherman, TX |
31/69 |
35/64 |
-5 |
Bluefield, WV |
39/59 |
43/56 |
-4 |
Beaumont, TX |
39/60 |
43/56 |
-4 |
Jackson, TN |
38/61 |
42/58 |
-4 |
Ft. Smith, AR |
34/64 |
37/62 |
-4 |
Charleston, WV |
43/55 |
46/53 |
-3 |
(In fact, because there aren't a lot more of them, I'll list every other media market where there was a negative Kerry-to-Obama trend, continuing on with #11: Knoxville, Tri-Cities, Lafayette, LA, Huntsville, Alexandria, LA, Shreveport, Wheeling, Tulsa, Youngstown, Chattanooga, Clarksburg, WV, Panama City, Pittsburgh, Nashville, Parkersburg, WV, Tucson, Paducah, and Tallahassee.)
Turning now to demographic information, this next table might actually be the most important, at least in terms of figuring how expensive it is to advertise in a particular market: the most populous and least populous markets, according to 2010 Census numbers. (If you're wondering how the seemingly arbitrary decision is made on whether or not a city gets its own market, it seems to boil down to whether that town has its own network affiliate or affiliates, rather than just rebroadcasting translators for other city's TV stations. So, yes, that means little Glendive, Montana makes the cut.)
Market |
Popu. |
Market |
Popu. |
New York, NY |
20,852,971 |
Glendive, MT |
10,145 |
Los Angeles, CA |
17,895,552 * |
North Platte, NE |
37,590 |
Chicago, IL |
9,642,027 |
Alpena, MI |
40,540 |
Philadelphia, PA |
7,980,444 |
Helena, MT |
69,007 |
San Francisco, CA |
7,303,245 |
Juneau, AK |
71,002 |
Dallas, TX |
7,020,482 |
Presque Isle, ME |
71,870 |
Washington, DC |
6,390,174 |
Zanesville, OH |
86,074 |
Atlanta, GA |
6,336,935 |
Victoria, TX |
86,793 |
Boston, MA |
6,322,433 |
Mankato, MN |
121,957 |
Houston, TX |
6,269,293 |
St. Joseph, MO |
127,329 |
(That asterisk is because the number actually reflects the sum of the Los Angeles and Palm Springs markets. Riverside County is the only very populous county in the nation that gets split between markets, so I don't give an accurate breakdown of how the population within that county actually splits... the Palm Springs market portion of Riverside Co. is probably only several hundred thousand, though.)
Turning to race and ethnicity data now, here are the markets with the highest and lowest percentage of non-Hispanic white residents. Note that there's a lot of correlation between the whitest, and least populous, markets:
Market |
White % |
Market |
White % |
Alpena, MI |
96.8 |
Laredo, TX |
3.5 |
Parkersburg, WV |
95.9 |
Harlingen, TX |
8.6 |
Clarksburg, WV |
95.3 |
El Paso, TX |
16.7 |
Bangor, ME |
95.2 |
Honolulu, HI |
22.7 |
Presque Isle, ME |
95.1 |
Yuma, AZ |
25.1 |
Charleston, WV |
94.7 |
Miami, FL |
27.7 |
Glendive, MT |
94.6 |
Greenwood, MS |
33.4 |
Wheeling, WV |
94.3 |
Los Angeles, CA |
33.6 |
Tri-Cities, TN |
94.3 |
Fresno, CA |
33.7 |
Portland, ME |
94.3 |
Corpus Christi, TX |
33.7 |
Moving on to the markets with the highest and lowest percentages of non-Hispanic African-American residents:
Market |
Black % |
Market |
Black % |
Greenwood, MS |
63.6 |
Laredo, TX |
0.2 |
Jackson, MS |
48.0 |
Alpena, MI |
0.2 |
Montgomery, AL |
43.8 |
Glendive, MT |
0.2 |
Memphis, TN |
41.8 |
Twin Falls, ID |
0.3 |
Meridian, MS |
40.3 |
Missoula, MT |
0.3 |
Macon, GA |
39.0 |
Butte, MT |
0.3 |
Albany, GA |
38.2 |
Helena, MT |
0.3 |
Columbia, SC |
37.9 |
Bend, OR |
0.3 |
Augusta, GA |
37.4 |
Harlingen, TX |
0.3 |
Columbus, GA |
37.1 |
Billings, MT |
0.4 |
The markets with the highest and lowest Hispanic percentages:
Market |
Hispanic % |
Market |
Hispanic % |
Laredo, TX |
95.6 |
Zanesville, OH |
0.8 |
Harlingen, TX |
90.0 |
Wheeling, WV |
0.8 |
El Paso, TX |
78.8 |
Parkersburg, WV |
0.8 |
Yuma, AZ |
74.3 |
Charleston, WV |
0.8 |
Corpus Christi, TX |
60.6 |
Bluefield, WV |
0.9 |
San Antonio, TX |
55.0 |
Presque Isle, ME |
0.9 |
Fresno, CA |
53.2 |
Bangor, ME |
1.0 |
Bakersfield, CA |
49.2 |
Clarksburg, WV |
1.0 |
Miami, FL |
48.1 |
Marquette, MI |
1.1 |
Odessa, TX |
48.0 |
Alpena, MI |
1.1 |
Finally, here are the markets with the highest and lowest Asian percentages:
Market |
Asian % |
Market |
Asian % |
Honolulu, HI |
37.7 |
Glendive, MT |
0.3 |
San Francisco, CA |
22.6 |
Zanesville, OH |
0.3 |
Los Angeles, CA |
12.1 |
Clarksburg, WV |
0.4 |
San Diego, CA |
10.6 |
Wheeling, WV |
0.4 |
Sacramento, CA |
10.3 |
Traverse City, MI |
0.4 |
New York, NY |
9.2 |
Alpena, MI |
0.4 |
Seattle, WA |
9.0 |
Presque Isle, ME |
0.4 |
Las Vegas, NV |
8.3 |
Meridian, MS |
0.4 |
Washington, DC |
8.2 |
Minot, ND |
0.5 |
Fresno, CA |
6.5 |
Bluefield, WV |
0.5 |
And last, let's talk about money. Again, this doesn't give a clear sense of which markets might charge more per eyeball because they contain more consumers with more discretionary income, but you can use this information to make an educated guess about it. There's no easy way to compile county numbers on per capita income or median household income into media-market-wide numbers. But what I did find available from the Census Bureau is the number of households in each county that fall into ten different income brackets; in a way, that's even better, because it gives a broader look at the income distribution in an area. (PCI by contrast only gives you one number; an area with an average PCI, for instance, might be very clustered around the middle with few extremes, or an area with a great deal of great inequality that happens to balance out in an average fashion.)
So, what I've done here is list the media markets that have the highest percentage of their households in each of the ten income brackets, which, to avoid blowing up your monitor, will take four different tables to show. For example, the Greenwood, MS, market, which is probably the most impoverished market in the nation, has the highest percentage of households in all three lowest brackets.
Market |
% in under
$10K bracket |
Market |
% in $10-15K
bracket |
Market |
% in $15-25K
bracket |
Greenwood, MS |
18.9 |
Greenwood, MS |
11.3 |
Greenwood, MS |
17.8 |
Harlingen, TX |
15.4 |
Harlingen, TX |
10.1 |
Alpena, MI |
17.5 |
Monroe, LA |
13.7 |
Bluefield, WV |
9.7 |
Bluefield, WV |
17.1 |
Meridian, MS |
13.7 |
Columbus, MS |
9.6 |
Meridian, MS |
16.8 |
Columbus, MS |
13.0 |
Eureka, CA |
9.5 |
Presque Isle, ME |
16.3 |
Gainesville, FL |
13.0 |
Jonesboro, AR |
9.3 |
Harlingen, TX |
16.1 |
Albany, GA |
12.9 |
Meridian, MS |
9.3 |
Clarksburg, WV |
16.1 |
Tallahassee, FL |
12.5 |
Presque Isle, ME |
9.2 |
Albany, GA |
16.0 |
Lexington, KY |
12.4 |
Monroe, LA |
9.1 |
Charleston, WV |
15.7 |
Hattiesburg, MS |
12.2 |
Tri-Cities, TN |
9.1 |
Columbus, MS |
15.7 |
A different set of markets start to emerge as we move into the markets with disproportionately high numbers of middle-income households:
Market |
% in $25-35K
bracket |
Market |
% in $35-50K
bracket |
Market |
% in $50-75K
bracket |
Twin Falls, ID |
14.8 |
Twin Falls, ID |
18.0 |
Salt Lake City, UT |
22.1 |
Springfield, MO |
14.2 |
Springfield, MO |
17.6 |
Green Bay, WI |
21.8 |
Ottumwa, IA |
14.0 |
Traverse City, MI |
17.5 |
Lima, OH |
21.8 |
Wheeling, WV |
13.9 |
Watertown, NY |
17.2 |
Bend, OR |
21.7 |
Joplin, MO |
13.9 |
Ft. Wayne, IN |
17.2 |
Ft. Wayne, IN |
21.6 |
Rapid City, SD |
13.7 |
St. Joseph, MO |
17.0 |
Sioux City, IA |
21.6 |
Alpena, MI |
13.6 |
Joplin, MO |
17.0 |
Boise, ID |
21.6 |
Traverse City, MO |
13.6 |
Zanesville, OH |
16.9 |
Mankato, MN |
21.6 |
Yuma, AZ |
13.6 |
Sioux City, IA |
16.8 |
Harrisburg, PA |
21.5 |
Myrtle Beach, SC |
13.5 |
Helena, MT |
16.8 |
Glendive, MT |
21.4 |
Interestingly, a lot of state capitols show up in the upper-middle range:
Market |
% in $75-100K
bracket |
Market |
% in $100-150K
bracket |
Juneau, AK |
15.4 |
Washington, DC |
19.2 |
Anchorage, AK |
15.3 |
Juneau, AK |
18.6 |
Honolulu, HI |
15.0 |
Anchorage, AK |
18.1 |
Fairbanks, AK |
15.0 |
San Francisco, CA |
17.6 |
Madison, WI |
14.8 |
Fairbanks, AK |
17.2 |
Minneapolis, MN |
14.6 |
Boston, MA |
17.2 |
Salt Lake City, UT |
14.5 |
Honolulu, HI |
17.2 |
Harrisburg, PA |
14.4 |
Baltimore, MD |
16.8 |
Hartford, CT |
14.3 |
Hartford, CT |
16.6 |
Helena, MT |
14.3 |
San Diego, CA |
15.6 |
And, finally, here's where you rabble-rousers can find the 1% (or technically, the 4.1% of households who fall in the $200K or more bracket):
Market |
% in $150-200K
bracket |
Market |
% in over
$200K bracket |
Washington, DC |
9.9 |
San Francisco, CA |
10.6 |
San Francisco, CA |
9.0 |
Washington, DC |
10.4 |
Boston, MA |
7.3 |
New York, NY |
8.7 |
New York, NY |
7.2 |
Boston, MA |
7.1 |
Baltimore, MD |
7.0 |
Baltimore, MD |
6.3 |
Anchorage, AK |
6.7 |
Monterey, CA |
6.2 |
San Diego, CA |
6.7 |
San Diego, CA |
6.2 |
Hartford, CT |
6.4 |
Los Angeles, CA |
6.2 |
Honolulu, HI |
6.4 |
Chicago, IL |
5.6 |
Los Angeles, CA |
6.1 |
Santa Barbara, CA |
5.6 |
If you're looking for the entire dataset -- and I mean entire, with all of this data at the county level -- it's available at Google Docs.