Bang-for-the-Buck Index: House Edition

Time for the thrilling conclusion to the Bang-for-the-Buck Index, begun yesterday with the Senate installment. Follow the link for full methodological nitty-gritty, but the main thing that you need to know is that this index shows which races are the cheapest media-wise (and thus where one netroots dollar gets stretched the furthest). This list covers all House races that Swing State Project projects as Dem pickup opportunities.

The middle column lists every media market that needs to be utilized in order to blanket the district, and the number next to each market is the number of thousands of TV households in that market. The more TV households, the more expensive the market. (When a market only grazes a small part of the district where there’s no major population center, I’ve deemed the market negligible, assuming that a smart media buyer wouldn’t use that market.) The number in the right column is the sum total of the thousands of TV households in all markets in the district, which provides a relative number that indicates how expensive a media campaign in that district is.

As you’ll see, there’s a huge amount of variation, depending on the number of ‘wasted eyeballs.’ The wasted eyeballs problem becomes huge in suburban districts in major metropolitan districts, where you may be paying to advertise to people in the adjacent 10 or 20 districts as well.

Let’s start with the cheap races:

District Markets Score
WY-AL Cheyenne (54)

Casper (52)

Denver (1,415 *)

Salt Lake City (811 *)

Rapid City (91 *)

Billings (103 *)

Idaho Falls (115 *)
106 *
AK-AL Anchorage (141)

Fairbanks (32)

Juneau (24)
197
LA-05 Monroe (174)

Alexandria (93)

Lake Charles (negligible)

Lafayette (negligible)
267
LA-07 Lafayette (220)

Lake Charles (94)
314
NE-02 Omaha (400) 400
AL-02 Montgomery (245)

Columbus GA (205)

Dothan (98)
548
SC-01 Charleston SC (284)

Myrtle Beach (266)
550
IA-04 Des Moines (414)

Rochester MN (143)

Cedar Rapids (negligible)

Sioux City (negligible)
557
LA-04 Shreveport (382)

Alexandria (93)

Lake Charles (94)
569
WV-02 Charleston WV (478)

Clarksburg (109)

Washington DC (2,253 *)
587 *
IN-03 South Bend (333)

Fort Wayne (271)
604
ID-01 Boise (230)

Spokane (390)
620
NV-03 Las Vegas (651) 651
NM-01 Albuquerque (654) 654
VA-02 Norfolk (705) 705
IL-18 Peoria (242)

Champaign (378)

Quincy (104)

Davenport (negligible)
724
NY-25 Rochester (385)

Syracuse (398)
783
OH-01 Cincinnati (880) 880
OH-02 Cincinnati (880)

Columbus OH (negligible)

Charleston WV (negligible)
880
OH-15 Columbus OH (891) 891
NV-02 Las Vegas (651)

Reno (255)

Salt Lake City (811 *)
906 *
SC-02 Columbia SC (373)

Augusta (246)

Savannah (296)
915
NM-02 Albuquerque (654)

El Paso (291)

Odessa (negligible)
945
KS-04 Wichita (447)

Tulsa (510)
957

The more expensive races are over the flip…

District Markets Score
KY-02 Louisville (643)

Evansville (289)

Bowling Green (75)
1,007
VA-05 Richmond (511)

Roanoke (440)

Charlottesville (70)

Raleigh (negligible)
1,021
CA-50 San Diego (1,026) 1,026
NY-26 Buffalo (644)

Rochester (385)
1,029
IN-04 Indianapolis (1,054)

Lafayette IN (63)
1,117
AL-03 Birmingham (717)

Montgomery (245)

Columbus GA (205)

Atlanta (negligible)
1,167
MO-06 Kansas City (904)

St. Joseph (46)

Columbia MO (168)

Ottumwa (51)

Omaha (negligible)
1,169
PA-18 Pittsburgh (1,170) 1,170
MD-01 Baltimore (1,089)

Salisbury (148)
1,237
FL-08 Orlando (1,346) 1,346
FL-24 Orlando (1,346) 1,346
OH-07 Columbus OH (891)

Dayton (514)
1,405
FL-18 Miami (1,523) 1,523
FL-21 Miami (1,523) 1,523
NY-29 Buffalo (644)

Rochester NY (385)

Syracuse (398)

Elmira (97)
1,524
OH-16 Cleveland (1,542) 1,542
MO-09 St. Louis (1,222)

Columbia MO (168)

Quincy (104)

Ottumwa (51)
1,545
PA-03 Pittsburgh (1,170)

Erie (159)

Youngstown (277)
1,606
MN-02 Minneapolis (1,653) 1,653
MN-03 Minneapolis (1,653) 1,653
MN-06 Minneapolis (1,653) 1,653
AZ-01 Phoenix (1,660)

Albuquerque (negligible)
1,660
AZ-03 Phoenix (1,660) 1,660
WA-08 Seattle (1,702) 1,702
FL-09 Tampa (1,710) 1,710
FL-10 Tampa (1,710) 1,710
CO-04 Denver (1,415)

Colorado Spgs. (315)
1,730
CA-04 Sacramento (1,346)

Chico (191)

Reno (255)
1,792
OH-14 Cleveland (1,542)

Youngstown (277)
1,819
NC-10 Charlotte (1,020)

Greenville SC (815)
1,835
MI-09 Detroit (1,936) 1,936
TX-07 Houston (1,939) 1,939
FL-25 Miami (1,523)

Ft. Myers (462)
1,985
FL-15 Orlando (1,346)

W. Palm Beach (752)
2,098
FL-13 Tampa (1,710)

Ft. Myers (462)
2,172
VA-10 Washington DC (2,253) 2,253
VA-11 Washington DC (2,253) 2,253
TX-10 Houston (1,939)

Austin (589)
2,528
NC-08 Charlotte (1,020)

Greensboro (652)

Raleigh (985)

Myrtle Beach (266)
2,923
PA-06 Philadelphia (2,926) 2,926
PA-15 Philadelphia (2,926) 2,926
MI-07 Detroit (1,936)

Toledo (427)

Lansing (257)

Grand Rapids (732)
3,352
IL-06 Chicago (3,431) 3,431
IL-10 Chicago (3,431) 3,431
IL-13 Chicago (3,431) 3,431
PA-05 Pittsburgh (1,170)

Buffalo (644)

Harrisburg (707)

Wilkes-Barre (589)

Erie (159)

Elmira (97)

Johnstown (295)
3,661
IL-11 Chicago (3,431)

Peoria (242)

Davenport (308)
3,981
CA-26 Los Angeles (5,536) 5,536
CA-46 Los Angeles (5,536) 5,536
CA-45 Los Angeles (5,536)

Palm Springs (143)
5,679
CT-04 New York (7,380)

Hartford (negligible)
7,380
NY-13 New York (7,380) 7,380
NJ-05 New York (7,380) 7,380
NJ-07 New York (7,380) 7,380
NJ-03 New York (7,380)

Philadelphia (2,926)
10,306
NJ-04 New York (7,380)

Philadelphia (2,926)
10,306

You may have noticed a few asterisked races; I’ll explain each one. WV-02 is partially covered by the Washington media market, which reaches into the tip of the panhandle (which is rapidly turning into DC exurbs). Advertising in DC is prohibitively expensive, so I’ve excluded it even though the panhandle is a populous part of the district. Like Manchester, New Hampshire (which we talked about yesteray), however, this is an unusual situation where there’s a single station nearby that’s considered to operate within the larger DC market, in this case in Hagerstown, Maryland. It’s likely that most of the WV-02 advertising targeting the panhandle would go through this one station.

NV-02 is partly covered by the Salt Lake City market (the easternmost three counties). This area contains fast-growing Elko, so it can’t be written off entirely, but again, it’s unlikely that any media strategy here would include SLC.

And finally, Wyoming is a particularly perplexing case. Using just the in-state markets in Cheyenne and Casper, it’s the cheapest district anywhere. However, these two markets cover only about 50% of the state’s population; the rest is out-of-state markets like Denver and SLC, so a comprehensive broadcast-TV strategy would shoot Wyoming into very expensive district territory. Most likely, the outlying portions of Wyoming are targeted purely through direct mail, AM radio, possibly cable systems, and as Gary Trauner adeptly showed last time, face-to-face contact.

You may have also noticed a number of predominantly rural districts that should theoretically be cheap but in fact are very expensive; MI-07 and NC-08 are key examples, each of which are kind of located between major cities and wind up biting a corner out of a bunch of different markets. Poor PA-05 is the perhaps the worst example; it doesn’t even have any TV stations in its boundaries, but it takes bites out of about 8 surrounding markets. Districts like these, again, are probably dealt with creatively, with buys in some TV markets and more focus on cable and other media.

The focus on cable, direct mail, and the like also probably becomes more important in the most expensive urban markets (New Jersey, anyone?) where even the best-financed House candidate isn’t going to be able to go on the air much. As I said yesterday, much of this is conjecture (and I certainly welcome comment from anyone with more experience with campaigning in any of these districts, or media buying in general); it’s just a rough guide to help netroots donors find races where their dollars might be used particularly efficiently.

The Bang-for-the-Buck Index

Over the last few months, I’ve come out with a variety of indexes that focused on the potential efficiencies of electing various new senators and representatives (for instance, in terms of the largest overall right-to-left shift, here and here). Part of this is to help netroots donors have some quantitative basis for seeing how their meager dollars might be best invested.

One thing that’s been missing from the equation, though, is any sense of how effectively those netroots dollars will be spent: not in terms of whether the campaign is likely to make good decisions (which is unknowable), but in terms of whether the media markets where an election takes place are a good buy. Think about how much further one netroots dollar goes in a race in Wyoming or Alaska, where there are few eyeballs to reach and the media markets are cheap, versus a race in New York or Texas. And yet, the result is the same: one more senator or representative.

Last week, Nate at 538.com had a few very interesting posts on this topic (creating the Return on Investment Index, and, then analyzing specific states’ media markets). Nate’s concept of efficiency turned largely on the idea of ‘wasted eyeballs,’ in other words, advertising in markets that bleed across state lines and where therefore ads run in front of non-targeted voters. However, his analysis was focused on the presidential race, where the concept of ‘wasted eyeballs’ may be overrated as a problem; after all, even if ads run in swing state markets bleed over into non-swing states, the voters in those states are voting in the same presidential election too. It may not maximize efficiency in the way a skilled media buyer would like, but it’s all part of a bigger whole.

We don’t have that luxury in Senate, and especially House races, though. Take Virginia as a case in point: Mark Warner needs to advertise in the Washington DC market in order to reach suburban northern Virginia voters, but that means paying top dollar in one of the nation’s most expensive media markets to tell millions of Marylanders to vote for him. (Of course, he can afford it, so don’t break out the tiny violin yet.) Kentucky may be even worse from a pure efficiency standpoint; none of its markets are brutally expensive, but blanketing all of Kentucky means advertising in Evansville, Cincinnati, and Nashville.

Over the flip, let’s look at all of the competitive Senate pickup opportunities. I’ll explain the methodology and the asterisked races after the table; for now, all you need to know is that the lower the number on the right is, the less expensive the race is.

State Markets Score
Alaska Anchorage (141)

Fairbanks (32)

Juneau (24)
197
Maine Portland (407)

Bangor (143)

Presque Isle (31)
581
Idaho Boise (230)

Idaho Falls (115)

Spokane (390)

Twin Falls (60)

Salt Lake City (negligible)
795
Nebraska Omaha (400)

Lincoln (274)

North Platte (15)

Sioux City (157)

Denver (1,415 *)

Cheyenne (negligible)

Rapid City (negligible)

Wichita (negligible)
846 *
New Mexico Albuquerque (654)

El Paso (291)

Amarillo (190)

Odessa (negligible)
1,135
Oregon Portland (1,100)

Eugene (229)

Medford (163)

Bend (54)

Yakima (negligible)

Boise (negligible)

Spokane (negligible)
1,545
Mississippi Jackson (328)

Memphis (658)

Biloxi (136)

Columbus (187)

Hattiesburg (105)

Meridian (71)

Greenwood (77)

New Orleans (672 *)
1,562 *
Kansas Kansas City (904)

Wichita (447)

Topeka (171)

Joplin (154)

Tulsa (negligible)

Lincoln (negligible)

St. Joseph (negligible)
1,676
Oklahoma Oklahoma City (665)

Tulsa (510)

Sherman (124)

Wichita Falls (155)

Ft. Smith (273)

Shreveport (negligible)

Joplin (negligible)

Amarillo (negligible)
1,727
Colorado Denver (1,415)

Colorado Spgs. (315)

Grand Junction (65)

Albuquerque (negligible)
1,795
Minnesota Minneapolis (1,653)

Duluth (169)

Rochester (143)

Fargo (234)

Mankato (51)

La Crosse (negligible)
2,250
New Hampshire Boston (2,375 *)

Burlington (326)

Portland (407)
3,108 *
Kentucky Louisville (643)

Lexington (479)

Cincinnati (880)

Evansville (289)

Bowling Green (75)

Paducah (383)

Charleston WV (478)

Nashville (928 *)

Knoxville (negligible)

Tri-Cities (negligible)
3,227 *
North Carolina Charlotte (1,020)

Raleigh (985)

Greensboro (652)

Greenville NC (270)

Wilmington (168)

Greenville SC (815)

Norfolk (negligible)

Myrtle Beach (negligible)

Atlanta (negligible)

Chattanooga (negligible)
3,910
Virginia Washington DC (2,253)

Norfolk (705)

Richmond (511)

Roanoke (440)

Tri-Cities (324)

Charlottesville (70)

Harrisonburg (86)

Bluefield (negligible)

Raleigh (negligible)

Greensboro (negligible)
4,389
Texas Dallas (2,336)

Houston (1,939)

San Antonio (760)

Austin (589)

Waco (311)

Harlingen (319)

Corpus Christi (192)

Laredo (64)

Beaumont (167)

Tyler (256)

Sherman (124)

Victoria (30)

Wichita Falls (155)

Abilene (113)

San Angelo (53)

Amarillo (190)

Lubbock (152)

Odessa (135)

El Paso (291)

Shreveport (negligible)
8,176

The middle column lists each media market that’s in the state in question. The number next to each media market is the number of thousands of TV households in that media market, according to most recent Nielsen numbers. (I deemed some markets ‘negligible,’ if they were out-of-state markets that barely spilled over the border and represented 2% or less of the state’s total population, thus unlikely to be part of a media buy.) The number on the right is simply the sum of all the TV households in the relevant markets, in other words, the number of households that need to be paid for in order to more or less blanket the state.

As you can see, there’s a pretty clear correlation between the expensiveness of a state and how populous it is. (As Nate points out, the cost per household may vary a bit from market to market, depending on the desirability of that market to advertisers. For the most part, though, the basic rule is that the more eyeballs you need to reach, the more dollars you’re going to pay. So the larger the number on the right is, the more expensive your race is going to be.) However, there can be some variations, depending on the ‘wasted eyeballs’ factor. States with nice clean media markets that correspond roughly to state borders are cheaper than some states that have smaller populations but more porous boundaries (for instance, Oregon is cheaper to blanket than Kansas, while Colorado is cheaper to blanket than Kentucky).

There are a few races that I asterisked; generally, it’s because of the presence of an out-of-state market that covers more than 2% of the state’s population but that’s also cripplingly expensive to compete in and that probably wouldn’t be part of an intelligent media buying strategy. Nebraska is a prime example: about 4% of the state’s population (most of the big empty western part) is served by the Denver market. But c’mon: you aren’t going to see Scott Kleeb TV ads running in Denver. Smart media buying would probably focus on AM radio or direct mail in that part of the state instead. (Adding Denver at 1,415 to the calculus boosts the net cost in Nebraska to 2,261.)

Likewise, a few counties in Mississippi (3% of the state’s population) are in the New Orleans market. (Adding New Orleans at 672 boosts the cost in Mississippi to 2,234, making it a much less attractive prospect. Mississippi also takes in the somewhat expensive Memphis market, but that covers 12% of the state’s voters and can’t safely be ignored.) Also, the Nashville market covers 5% of Kentucky’s population. Bruce Lunsford can pay for that if he wants to, but adding Nashville at 928 boosts the already high costs in Kentucky up to 4,155.

Finally, there’s the question of New Hampshire. The bottom half of the state is considered part of the Boston market, but there is one affiliate based in Manchester that is considered to operate within the larger Boston market. (As Nate points out, it may owe its entire existence to New Hampshire’s weird role in the presidential race and the targeted ad blitz that results.) Not really knowing how that shakes out in terms of ad rates, I’m leaving New Hampshire as is, but figure that the actual costs in New Hampshire are probably lower. [Update: According to DavidNYC, in 2006, the House candidates in NH mostly focused on the Manchester affiliate and steered clear of Boston in general, although the DCCC did a whole-Boston-market moneybomb right before the election.]

And of course, there are the usual caveats that TV and radio advertising are only a portion of a sane advertising strategy, which includes everything from internet and direct mail down to the totally unglamorous world of yard signs and stickers. This is only a rough guide to give you an idea of how expensive a particular race may be, and how far your dollar might go toward making a difference.

I’ll take a look at the House races using the same method tomorrow.