The Facts about Television Advertising and the McCain-Feingold Bill

The Facts about Television Advertising and the McCain-Feingold BillAuthor(s): Jonathan Krasno and Kenneth GoldsteinSource: PS: Political Science and Politics, Vol. 35, No. 2 (Jun., 2002), pp. 207-212Published by: American Political Science AssociationStable URL: http://www.jstor.org/stable/1554720 .

Accessed: 14/06/2014 08:27

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

American Political Science Association is collaborating with JSTOR to digitize, preserve and extend access toPS: Political Science and Politics.

http://www.jstor.org

This content downloaded from 62.122.72.154 on Sat, 14 Jun 2014 08:27:26 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=apsa

http://www.jstor.org/stable/1554720?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


m mm Il VA FEATURES

The Facts about Television Advertising and the McCain-Feingold Bill*

Jonathan Krasno, Yale University Kenneth Goldstein, University of Wisconsin, Madison

The McCain-Feingold bill and its House counterpart, sponsored by

Representatives Shays and Meehan, are universally regarded as the most signifi- cant campaign finance legislation under serious consideration by Congress in a generation-perhaps since the 1974 amendments to the Federal Election Campaign Act (FECA). This legislation would not expand on the 1974 reforms but instead restore them by regulating the two mechanisms that have developed in the intervening decades to cir- cumvent FECA, so-called "soft money" and "issue advocacy."' Together and separately soft money and issue advocacy have become an enormous part of many federal campaigns, in some cases even eclipsing the efforts of candidates operating under FECA's rules.

That popularity, naturally, has created a powerful group of donors and recipients who have exploited these loopholes and now oppose any attempt to close them, even as a growing number of contributors have begun to complain of the relentless pressure to give money. These political forces, coupled with the putative relationship between soft money, issue advocacy, and several core constitutional values, have made McCain-Feingold among the most controversial bills facing Congress.

Jonathan Krasno is a visiting fellow at Yale University's Institute of Social and Policy Studies. He is author of Challengers, Com- petition, and Reelection (Yale University Press, 1994) and Buying Time (Brennan Center 2000), as well as articles on congressional elections and campaign finance. He can be reached at [email protected].

Kenneth Goldstein is an assistant professor in the department of political science, University of Wisconsin, Madison. He is the author of Interest Groups, Lobbying and Participation in America (Cambridge Uni- versity Press, 1999). His research focuses on political advertising, interest groups, and survey methodology. He can be reached at goldstei@polisci. wisc.edu.

This paper uses a unique source of data about television commercials to examine some of the most important issues connected to this proposal. It is appropri- ate that we focus on television advertising since it is the largest-and most discussed-single category of expenditures by candidates, parties and interest groups in federal elections.2 McCain-Feingold's chief impact would surely be seen on the nation's airwaves, on the thousands of issue ads paid for with soft money. Indeed, many of the arguments for and against McCain-Feingold are rooted in different interpretations of those very ads.

For its critics, the huge outlay on issue ads is a dangerous scam perpetrated on democracy, a scam predicated on twin falsehoods that issue ads promote issues and soft money builds parties. For its defenders, the spending on issue advertising is a sign of democracy's vitality and any attempt to limit issue ads or soft money is inherently ham-handed and dangerous. Fortunately, many of these claims are empirical questions; given the proper data they can be carefully dissected and weighed. That is precisely what we do, using an extensive and comprehensive data set on television advertising to ex- plore some of the core assumptions in- voked by proponents and opponents of McCain-Feingold.

Monitoring the Airwaves The sheer amount of television

advertising-on approximately 1300 stations in the nation's 210 media markets over the 15 or 16 most popular hours in the broadcast day-makes commercials extremely difficult to study. Fortunately, using satellite track- ing technology first developed by the U.S. Navy to detect Soviet submarines, the Campaign Media Analysis Group (CMAG) is able to gather information about the content, targeting and timing of each ad aired. CMAG tracks commercials by candidates, parties, and interest groups in the nation's top 75

media markets. Together these markets reach approximately 80% of house- holds in the U.S. CMAG's technology recognizes the seams in programming where commercials appear, creates a unique digital fingerprint of each ad aired, then downloads a version of each ad detected along with the exact time and station on which it appeared. The company later adds estimates of the average cost of an ad shown in the time period.

With funding from The Pew Charita- ble Trusts, CMAG's data for 1998 and 2000 were purchased. These data are literally a minute-by-minute view of political advertising across the country-along with a "storyboard" like the one we cite, a frame of video every 4-5 seconds plus full text of audio, for each ad detected during these two election cycles. The storyboards were then examined by teams of graduate and un- dergraduate students at the University of Wisconsin (2000) and Arizona State University (1998), who coded the content of each commercial.3

Some of the questions coders an- swered-such as whether an ad mentioned a candidate for office by name or urged viewers to "vote for" or "defeat" a particular candidate-were objective. Others were subjective. These included items asking coders to assess the purpose (to support a particular candidate or express a view on an issue) and tone (promote, attack, or contrast) of an ad. Both types of questions elicited nearly identical responses from different students who viewed the same ad, indicat- ing a reassuring degree of intercoder reliability.4 In addition, we also took special care to examine the disclaimer in each commercial, the written portion appearing usually at the end of each commercial noting its sponsor ("Paid for by . . ."), where possible.5 From this we were able to determine whether an ad was sponsored by a candidate, party, or interest group, and, if paid for by a party or group, whether it was an issue ad or not.

PSOnline www.apsanet.org 207



of how much was spent on issue ads by the parties and their allies over the last two cycles.

Figure 1 breaks down the issue ads in Table 1 by party, showing the total number run by various Democratic and Republican party committees and their allies.7 While Republicans had a slim advantage in issue ads in 1998, Democ- rats claimed a small lead in 2000. This modest reversal illustrates the unpre- dictability of soft money. Since contri- butions (to either parties or interest groups) for issue ads are unlimited, the generosity of a relatively small number of well-heeled donors may shift the tide. But equally striking was the near equality between the parties. Total soft money spending on television advertising for the Democrats and Republicans was separated by no more than $5 mil- lion in either year, a relatively small amount among the hundreds of millions spent on political advertising in both years.8 That is not to say, of course, that no candidates would have been particularly helped or hurt had McCain- Feingold been in effect earlier, only that the Democrats' and Republicans' gains and losses come fairly close to balanc- ing out across the country.9

Regulating Issue Advocacy The working definition of issue advo-

cacy comes from a footnote in the Supreme Court's seminal decision in Buckley v. Valeo (1976) that limited FECA's impact by defining campaign

WW". wwv w *.y z. - #w^s #^ lTE ^z[|

our nehbers' rnontt lind our that to szopW 11e wen ood up to hi5 1riends' rny, they just want the own p - y to rOuild Itw will lo spend Ihat meney hollwwor Ihqt wanl to"

An the zht dIt*lon. pun]: SUnds on olher pragr XWe_nl:

A "^p would lep CongfQss *l ff} he's a we st wr. An

from speffdJng sl securlly ho"2nt

ai sncefe person.*' {finnouncofJ: f!4llG h?k Nn aI S1 Sm to kee p rduilding *o?isl sucurtqf

ev |*w u4esnw.

Ad Games. Among the 2000 advertisements and storyboards reviewed by the authors was this advertisement by the Republican Party of Wisconsin that proised its 1998 Senate nominee, Mark Neumann.

Coders ended up examining approximately two thousand different federal ads (eliminating ads referring to state and local candidates or ballot proposi- tions) in 1998 and nearly three thousand ads in 2000. As Table 1 shows, these ads fell into different campaign-finance categories and appeared on the air hundreds of thousands of times. Most of the astonishing growth from 1998 to 2000, of course, is attributable to the presidential election, but the number of ads in congressional elections also rose in this two-year period from 293,942 to 477,190 and expenditures nearly dou- bled. House and Senate candidates ran more ads in 2000 than in 1998 (315,421 versus 235,516 in the earlier year), but most of the upsurge in these races came from parties and interest groups (58,426 ads to 159,511 in 2000).6

Whose Ox is Gored? The first question the professional

politicians in Congress will ask about McCain-Feingold is, "Who will it af- fect?" Such questions are always per- ilous since political operatives will un- doubtedly try to adapt to any new regulations, searching for new loopholes

208

to exploit. In which direction their search will eventually take them is at best an educated guess. What is more than guesswork, however, is the matter

PS June 2002



FIGURE 1 Soft Money Advertising by Parties and Interest Groups

1998 Pro-Dem

1998 Pro-GOP

2000 Pro-Dem

2000 Pro-OOP.

U ParUes OGroups

I I E U~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~rr----???-?I I

" .

.: . - .... .

giooo 100000

Number of ads

150000 - 200000

Furthermore, the issues raised in commercials by candidates and in issue ads were virtually identical. Table 2 lists the top five themes appearing in both types of ads in 1998 and 2000. Although occasional varia- tions appeared, the overwhelming impression is that issue ads mimicked the commercials that candidates ran. This may be mere coincidence, but it is a suggestive one. At very least, it con- tradicts the argument that issue ads by parties and interest groups introduce policy matters into the political arena that are otherwise ignored. The truth is that the issues emphasized by candidates were generally the only topics addressed by any advertiser, particularly in the final hectic weeks of the campaign. '

0

communications as those "expressly ad- vocating" the election or defeat of a particular candidate by using words like "elect," "defeat," or "support." The purpose behind the footnote was to protect speech about "issues"-lobbying on bills before Congress, pronouncements or de- bate over public policy-from the finan- cial regulations affecting partisan electioneering. The need to distinguish the two is obvious, but whether use of specific words of express advocacy (now widely known as "magic words") is an effective way to do so is less clear.

We sought to evaluate this standard by looking at ads purchased by candidates' campaigns. Candidates are a per- fect test case since the purpose of their advertising is so obviously electioneering that the magic words test does not apply to them. Thus, candidates must live with FECA whether or not they use magic words. That might lead one to assume that candidate ads unabashedly urge viewers to vote for one person or defeat another, but it turns out that such direct advocacy is exceedingly rare. In 2000 just under 10% of the nearly 432,190 ads paid for by federal candidates di- rectly urged viewers to support or oppose a particular candidate or used a slogan like "Jones for Congress," the full list of magic words in Buckley. Earlier we found that just 4% of 235,000 candidate ads in 1998 used any of the verbs of express advocacy; 96% did not ask viewers to vote for or against any candidate. Any device that fails to detect what it was designed to find nine times out of 10 is clearly a flop. The magic words test simply does not work.

There is also the matter of timing. If issue ads were intended only to pro- nounce on important policy matters, then we would expect to see them spaced throughout the year or concen- trated in periods when Congress is most active. As Figure 2 demonstrates, however, that was far from the case. While in both 1998 and 2000 members of Congress cast a steady stream of votes and a series of what Congressional Quarterly labels as "key votes" throughout the year, the greatest deluge of issue ads began appearing after Labor Day (about week 36).11 Indeed even the most casual inspection of the number of issue ads that appeared each week indicates that this line is much more closely related to the activity of candidates, not the activity of Congress. This relationship between advertisers and candidates, repeated over two years, is far too strong to be coincidental. There is no doubt that issue ads were largely in- spired by the same cause that motivates candidates, the slow approach of Election Day.

Despite the overwhelming evidence that the vast majority of issue ads were a form of electioneering, there were commercials in each year that our coders found to be genuine discussions of policy matters (22% of issue ads in 1998, 17% in 2000). Would the definition of electioneering created by McCain-Feingold-any ad mentioning a federal candidate by name in his or her district within 30 days of the primary or 60 days of the general election-inad- vertently capture many of these commercials? We addressed this question by comparing the issue ads that would have been classified as electioneering under McCain-Feingold to the coders' subjective assessment of the purpose of

PSOnline www.apsanet.org

--

?r??r?;::I?;??;?;?;?r

209



each ad. In 1998 just 7% of issue ads that we rated as presentations of policy matters appeared after Labor Day and mentioned a federal candidate. In 2000, that figure was less than 1%. Critics may argue that the chance of uninten- tionally classifying 7%, or even 1%, of genuine issue ads as electioneering makes this bill overly broad. In contrast, these percentages strike us as fairly modest, evidence that McCain-Feingold is finely calibrated. The 7% of genuine ads that would have been affected by McCain-Feingold in 1998 and 1% in 2000 are actually composed of just a handful of commercials-two in 1998 and three in 2000-that received varying degrees of play. Our examination suggests that these errors may be reduced with some small additions to the bill.12

Party Soft Money Just as the rules on issue advocacy

are intended to safeguard free speech,

soft money is also intended to achieve a worthy goal, in this case to strengthen political parties. Parties are a frequently underappreciated fact of political life in the U.S. Political scientists have sought ways to buttress them for years, to augment their ability to communicate with and mobilize the public, and to magnify their impact as political symbols.

The most obvious place to start as- sessing the value of parties' advertising is with a simple objective question: Does the ad mention either political party by name? It is hard to imagine how a commercial might strengthen a party if it neglects to praise its sponsors or at least malign the opposition. Yet, party ads are remarkably shy about saying anything about "Democrats" or "Republicans"- just 15% of party ads in 1998 and 7% in 2000 mentioned either political party by name. By contrast, 95% of these ads in 1998 and 99% in 2000 did name a particular candidate. It seems fairly clear that these ads do far more to promote

FIGURE 2 The Relationship between TV Ads and the Congressional Agenda

I F.;' . " -` '' o g aI ( m S I5 "Mk. ' of thte" -)i, (

600

500

J 400

5 300

200

.100

0

the fortunes of individual candidates than the fortunes of their sponsors. Indeed, the sample storyboard for this article was an advertisement by the Republican Party of Wisconsin that praised its 1998 Senate nominee, Mark Neumann, for "standing up to his own party." That ad may have gone farther to distinguish between candidates and parties than most of the others produced by parties, but it illustrates that parties themselves take a far backseat to candidates in Democratic and Republican issue ads.

A piece of supporting evidence for this conclusion comes from the per- ceived negativity of each ad. Coders found ads by parties to be much more likely than ads by candidates to be pure attack ads (60% in 1998, 42% in 2000).13 To be clear, we make no claim that attack advertising is somehow better or worse than other forms. Still, there is little hope that this flood of negative commercials magically strengthens either party.

Finally, some defenders of party soft money argue that these commercials help provide vital information to voters that they would not otherwise receive. It is obviously debatable whether any particular ad conveys much information to viewers. If, for the sake of argument, we accept this assertion and assume that party ads help educate voters, then their value is related to how they are allocated. Do party ads appear for candidates about whom little is known or in otherwise neglected districts and media markets? If the answer is yes, party ads may play an important role in informing the public. If no, then their potential value to voter edu- cation cannot be great.

Two facts suggest the latter conclusion. First is the sheer concentration of party ads in just a handful of races. In both 1998 and 2000 more than half of party ads in the top 75 media markets ran in just three Senate and a dozen House contests, even though the CMAG system tracks advertising in scores of states and hundreds of districts. Second is the close relationship between volume of commercials by candidates and parties. While there are a few exceptions- the Republican National Committee sponsored all of the pro-Bush advertising in California and neither party ran commercials in New York after the two Senate candidates agreed to forgo soft money-in the vast majority of cases parties were most active in the very places already saturated by candidate ads. Indeed, the level of candidate activity remains one of the best predictors of the level of party activity in a particular race. It may make political sense

PS June 2002

*- s = Yf a :'S st ; "s? .

'W,Jk of'th"f'YiMr: (2(0)

210



for parties to concentrate their efforts in a small number of the most hotly con- tested races-assuming diminished returns are not a concern-but this allo- cation strategy undercuts their defenders' claims that party advertisements serve an important civic role. It also undercuts the more plausible claim that parties' ads do much to increase the number of competitive House and Sen- ate elections since parties were most active in races already made competitive by the candidates.

Conclusion Our examination of television com-

mercials in 1998 and 2000 demonstrates that the way in which our campaign finance laws have been interpreted and

applied is unmistakably flawed. The magic words test supposed to distinguish issue advocacy from electioneering is a complete failure. The rules allowing parties to collect unlimited amounts of soft money to build stronger parties have instead allowed candidates to use parties as a way to raise unlimited funds and help themselves. Political scientists, of course, will

recognize that an important way for

parties to strengthen themselves is to

augment their share of officeholders.

Notes *An earlier version of this essay was distrib-

uted on Capitol Hill before both the Senate and House campaign-finance debates, and was re- ferred to in those debates.

1. These terms are familiar to most newspa- per readers, if not always understood. Soft money refers to money raised by parties and interest groups that is not used to support the election or defeat of a federal candidate-such as a donation to build a party headquarters-and thus avoids the source and size restrictions in the Federal Election Campaign Act. Over time, parties and groups have found ways to use soft money for electioneering, thus fueling the controversial rise in soft money receipts. The most important of these ways is issue advocacy, advertisements ostensibly addressing an issue with- out explicitly exhorting viewers to support or oppose a particular candidate. See below.

2. See a series of reports about "outside money" by Professor David Magleby of Brigham Young University. <www.byu.edu/ outsidemoney/>.

3. We thank particularly Daniel Seltz (1998), Lee Bradford (1998), Travis Ridout (2000), and Mike Franz (2000) for their work with the storyboards.

4. To test reliability, we had a sample of 250 ads recoded by three different coders. Agree- ment among the coders was extremely high, particularly on the items discussed here. Coders differed about the purpose-"In your opinion, is the purpose of this ad to provide information about or urge action on a bill or issue, or is it

Unfortunately, the legislators, regulators, and judges who created the category of soft money saw it as a way for parties to build more muscular organizations and perhaps buttress their symbolic appeal. Our examination of parties' ads in the last two election cycles shows that these hopes have been abandoned in favor of aiding a relatively small group of candidates. Any contention that most issue ads are motivated by issues or that most soft money builds political parties must ignore a veritable mountain of conflicting evidence. Such claims are completely unsustainable.

Whether that conclusion should trans- late automatically into support for McCain-Feingold is a different matter. These decisions inevitably involve a number of factors, starting with the judgment whether this bill is the best re- sponse to the manifest weaknesses of our campaign finance laws. We cannot be sure that it is, but our analysis suggests two important facts in its favor. First, the experience of the last two elections suggests that neither the Democratic nor Re- publican party would be disproportion- ately harmed by the ban on soft money or the stricter definition of issue advocacy that are the core of McCain-Feingold. In- deed, neither party stands to gain or lose much against their counterparts, at least

to generate support or opposition to a particular candidate?"--of only one ad. They were also close to unanimous in their responses to the question about the tone of ads: "In your judge- ment, is the primary purpose of the ad to promote a specific candidate ('In his distinguished career, Senator Jones has brought millions of dollars home. We need Jones.'), to attack a candidate ('In his long years in Washington, Senator Jones has raised your taxes over and over. We can't afford 6 more years of Jones.'), or to contrast the candidates ('While Senator Jones has been raising your taxes, Representa- tive Smith has been cutting them.')?" Coders disagreed about whether only two of the commercials were attack or contrast ads, and reached identical conclusions on all of the positive spots. This consensus is perhaps more re- markable because we chose not to embark on any training program for coders, preferring instead that they use their common sense to give us a better feel of how the average viewer would characterize these ads.

5. This language also allowed us to distinguish between issue ads and the other sorts of commercials that parties and interest groups may run, so-called "coordinated" and "inde- pendent expenditures." Because of myriad practical advantages, virtually all commercials by parties and groups were issue ads. A fair number (over 10%) of disclaimers were either missing or illegible in each year, perhaps appearing against a dark background or in the seconds between frames pictured in the story-

not on television. Of course, both sides will try to adapt to the law in ways that may make its eventual impact less even- handed. For starters, we expect Republi- cans to seek ways to maximize their advantage in hard money and Democrats to rely even more heavily on the superior organizations of their allies in labor and an array of membership groups. At pres- ent, however, McCain-Feingold's immedi- ate cost to both sides in television advertising is almost equal.

Second, we found no evidence that the new dividing line between issue advocacy and electioneering in McCain- Feingold is overly broad and would af- fect many commercials that we found to be genuine attempts to advocate issues, not candidates. Some critics will surely complain that we have no objective standards for determining which commercials are genuine issue advocacy, but that is untrue. The standards offered in McCain-Feingold are objective. That they perform so well against the subjective judgment of our coders, each of whom examined hundreds of ads, is extremely reassuring. We are always eager to consider improvements, but there is no reason not to conclude that the definition of electioneering in McCain- Feingold is, at the very least, an excel- lent start.

boards. In those cases we relied on the text of the ad itself and media reports to categorize the commercials.

6. Though the House/Senate ads and party/ interest group ads for 2000 (315,421 and 159,511, respectively) do not equal the total number of ads (477,190), 2,258 ads were un- clear as to who sponsored them.

7. Each ad judged as electioneering was also coded according to which party's candidate was favored.

8. Strictly speaking, a mixture of hard and soft money is used by parties in some issue ads. The situation is further complicated by maze of intraparty transfers between national and state organizations.

9. The nationwide balance between pro- Democratic and pro-Republican ads by groups and parties is not repeated in every district, state or media market. In cases such as the RNC's late incursion into California on behalf of George Bush, political and local considera- tions influence parties and groups to advertise in areas largely ignored by their counterparts. Furthermore, local advertising strategies may occasionally result in periods where a candidate's allies pay more for fewer ads than party organizations or groups on the other side. This illustrates one of the most important facts about TV advertising, its variation of cost during different times in the broadcast day and during different periods of the year. Thus, while accurate and intuitive, spots may give a misleading impression of campaign activity

PSOnline www.apsanet.org 211



since some ads are much cheaper to purchase than others. Dollar estimates, however, are only estimates of the average cost of a time slot and certainly undervalue ads aired near election day. All of the results reported in this paper hold for both units of analysis.

10. Earlier in the year, issue ads often cov- ered topics distinct from the advertising that candidates later ran.

11. Most of the key votes in each year occurred well before Labor Day, yet most of the issue advertising appeared afterward.

12. See "Five New Ideas to Deal With the Problems Posed by Campaign Appeals Mas- querading as Issue Advocacy," Brennan Center for Justice (2000) (<www.buyingtime.org>), for several suggestions about amending this proposal to reduce further its impact on these genuine

issue ads. In addition, we think it is likely that sincere lobbying efforts could find a way to broadcast an appeal that avoids McCain-Feingold by omitting the name of specific members of Congress.

13. Parties were the most negative advertiser in 1998, interest groups in 2000. Candidates in both years tended to emphasize positive information about themselves.

PS June 2002 212



Documents

The Facts about Television Advertising and the McCain-Feingold Bill