Do Political Campaign Ads Actually Work? Research Suggests Not Really

Ever since Spiro Agnew’s 1970 speech blasting “nattering nabobs of negativism” in the media, attacking the political neutrality of the mainstream press has been a frequent activity of conservatives. By and large, however, the right-wing response to the media unfairness has been merely to complain about the situation rather than do something about it.

Instead of buying, boosting, or starting mainstream media outlets, the primary response of conservatives to counter perceived bias against them in the press has been to purchase television, radio, and direct mail advertising. This has primarily been at the behest of failed political consultants who tell candidates and interest groups that they can get the perfect “messaging” which will be able to magically neutralize months of negative press and also cancel out the themes being put out by the opposition.

To such people, liberal media bias exists but it is something that they in their great wisdom can overcome.

Contrary to the consultants’ claims, however, empirical research conducted by political scientists shows that such advertising campaigns are not nearly as effective as many donors have been led to believe—especially in presidential campaigns.

This year is a unique one in political history in that it will mark the first time ever that many high-dollar GOP donors have decided to opt out of a presidential election. If they had bothered to look at the political science scholarship about televised campaign advertising, they’d have made this decision long ago.

There are five reasons why:

1. Research shows that almost half of all political ads are completely ignored and those that are watched do not change opinions.

People have not only become jaded about ads generally, they are increasingly able to avoid seeing them at all thanks to the proliferation of television channels, digital video recorders and web ad blocking technologies.

According to research done at the University of California–Los Angeles and Stanford University, television viewers tune out almost half, 42 percent, of all campaign commercials—they never even bother to watch them at all.

A survey commissioned for the Republican-allied Crossroads Generation group focusing on younger voters found that only about 25 percent of respondents said they used TV ads as a source of political news. This was lower than the number of young people who said they got their news from comedy programs like The Daily Show in a separate survey commissioned by the College Republican National Committee.

According to the CRNC report, 40 percent of respondents said they used some form of digital video recorder like TiVo to watch television. A 2013 study by DVR manufacturer Motorola found that 74 percent of American DVR users say they purchased their devices in order to fast-forward commercials. Needless to say, people who want to pay money for the ability to skip commercials are not interested in watching them.

But even under ideal conditions outside of the political world, television advertising has been found to have very limited positive effects.

The perfect illustration of this is the Super Bowl television spots that have become famous in their own right. But even these ads have been shown to be overwhelmingly ineffective.

This is notable since the championship football game is probably the only time that at least some people look forward to watching ads. In fact, 25 percent of Americans say they view the ads as the most important part of the show.

Part of the reason that people actually like Super Bowl commercials is that they frequently use humor and risqué themes that no politician would ever dare deploy. Even under these ideal circumstances, the spots have very little effect. According to a two-year study by the advertising research firm Communicus, most Super Bowl ads do not have any positive effect on persuading potential customers to buy the product or service being touted.

Of 40 different advertisements studied, 15 of them actually made poll respondents less likely to purchase the product or service being hawked. By contrast, only 13 companies saw an increase in persuasion. The rest saw no significant effect of any kind. That means that 68 percent of Super Bowl advertisements either hurt the brand paying for it or do nothing at all to help it.

Considering that most people hate them, political spots have a much more difficult task than Super Bowl advertising. And the research confirms what you would expect in light of the circumstances. According to the academic research, politician ads have almost no effect whatsoever at the presidential level.

According to the UCLA-Stanford study mentioned earlier, the people most likely to sit through television political ads are what the authors termed “low engagement” individuals, people who don’t really care about politics and don’t vote. The next group most likely to tune in are “high engagement” voters, political junkies who already have their minds made up.

The most highly prized group—the “medium engagement” voters that campaign consultants brag about their ability to reach since they generally vote and are most open to changing their minds—are the people most likely to tune out political ads, especially as the number of those ads on TV increases. The more political advertising these people see, the more they are likely to fast-forward or change the channel.

The study’s authors summarized their findings thusly: “Since the low engagement voters probably will not turn out to vote anyway and the high engagement voters likely made up their minds long ago, it is bad news for the candidates that the group they most need to sway is the group most likely to censor them.”

Of the few ads which are actually watched and paid attention to, most of them don’t really do much persuading. Instead, they simply make partisans of each side feel good about themselves. Television advertising is particularly ineffective when an incumbent whom voters already know and have likely formed their own opinions about is on the ballot.

This was well documented in a 2010 study by political scientists Michael Franz and Travis Ridout who found that advertising in the 2004 campaign had very little effect on people’s willingness to vote for either Republican George W. Bush (the incumbent) or Democrat John Kerry.

The study found that Kerry and his allies “had consistent advantages over the Bush camp in spots aired.” In total, Bush and his allies aired 408,604 spots either promoting him or attacking Kerry, far fewer than the 605,533 which backed Kerry.

Despite the fact that voters saw far more anti-Bush and pro-Kerry ads, the study found that they had almost no effect on voters. In counties where Kerry had a 1,000-ad advantage over Bush, this yielded a vote increase of less than 1 percent. It was not even a half a percent. It was 0.19 percent. And that was in counties which were not “battleground” ones where voters were being bombarded constantly by advertisements.

When Franz and Ridout looked at all counties, they found that a 1,000-ad advantage yielded Kerry an even smaller increase: 0.12 percent. When they looked at just September and October, they found that pro-Kerry ads were even less effective with 1,000 more ads netting him just 0.083 percent more votes.

With no incumbent in the race in 2008, the authors found ads were slightly more effective. Across all counties, they found that a 1,000 spot advantage for Democrat Barack Obama or his allies gained him 0.551 points. They were at their most useful in October in non-battleground counties where a 1,000-spot advantage helped Obama gain 1.081 percent more votes. Running more ads in these counties did not help Obama persuade anyone, however; instead, they merely helped him pad his vote total by motivating his base.

In 2012, with Obama running as the incumbent, advertising once again became less effective. Georgetown University professor Daniel Hopkins applied this same methodology to the 2012 race and found a similarly low effect on Obama’s vote totals. A 1,000-ad disparity in favor of the president resulted in a vote gain of just 0.14 percent for him.

A similar dynamic is in place for congressional elections as well. Campaign spending simply is not as effective as donors and politicos wish it were. According to an analysis of 2014 Senate races by Alan I. Abramowitz of Emory University, spending by both political parties and outside groups “had no discernible impact on the Democratic candidate’s margin in these contests.” Instead, the biggest factors at work were whether an incumbent was in the race and the way the state in question had voted in for president in 2012.

The just-concluded Republican presidential primary also showed that spending big has basically nothing to do with actually winning. Despite spending far less than his rivals, businessman Donald Trump was able to clinch the nomination with a vote percentage just slightly lower than Mitt Romney did in 2012.

Among the several candidates for president, Trump and his allies spent around $65 million compared to $160 million for Jeb Bush (who dropped out early after losing big in South Carolina), $140 million for Ted Cruz, $111 million for Marco Rubio, and $75 million for Ben Carson.

2. The effects of campaign ads are easily countered

When advertising does work, its effects are comparatively small and usually only in markets the other candidate is basically ignoring. In some cases, an advertising disparity can be overcome simply by having the candidate visit an area being blitzed by the opponent and receiving free (and usually positive) local media coverage of the event.

The study by Franz and Ridout mentioned earlier noted that the coverage could have slightly or very positive effects:

Bush and Kerry visits had a small and insignificant influence on the vote in 2004, though it was in the expected direction, but both Obama and McCain visits were correlated with a significant shift in the vote in 2008 (at about twice the size of the shifts that resulted from Kerry and Bush visits). The effect in nonbattleground states dissipates for Obama, but McCain visits net stronger support for counties in nonbattleground states. This spillover effect is still important in nonbattleground contexts. Because people in counties in spillover markets watch major network affiliates in the battleground states, they are exposed to local news coverage of the candidate visits in those states and are presumably susceptible to the positive press that these visits might engender. This was largely true for McCain in 2008.

Franz and Ridout’s research also indicates that a campaign’s reliance on outside groups to provide funding for ads generally support of it can dilute the overall message and thus make the expenditures less effective. This dynamic was at work in the 2004 campaign against John Kerry, in 2008 against John McCain and in the 2012 campaign against Mitt Romney. By contrast, winning candidates George W. Bush and Barack Obama both controlled their messages and their fund-raising.

Some ads can indeed be useful tools to increase turnout but to suppose that spending vast sums of money on “negative” ads designed to get around liberal media bias is sound strategy is simply incorrect.

Trying to give potential voters unflattering information about the opposing candidate through advertising does not appear to work any better than more positive messages, in large part because negative campaigns tend to also stir up partisans of the opposition. In fact, a meta-study of negative advertising published in The Journal of Politics found that candidates who engaged in it actually saw a slight decline in their own support.

The study’s authors cautioned against saying that negative campaigning is generally harmful to the politician engaging in it but what they did feel comfortable saying that it was clear to them that negative campaigning is not effective:

We uncovered 43 relevant findings in the literature, 27 involving intended vote choices and 16 involving reported vote choices or official vote totals. […] Only 12 of these outcomes are positive from the attacker’s standpoint. Only five of these 12 are at all appreciable, and they are counterbalanced by five negative effects of similar magnitude. The remaining effects are all small. The overall unadjusted effect of negative campaigns on vote choice indicates a modest disadvantage to the attacker, but that effect vanished when we adjusted the effect sizes. It bears mentioning that the two largest negative effects and all five of the positive ones come from experimental studies in which the dependent variable is vote intention; that particular design evidently produces more volatile outcomes. The broader message, though, is that the research literature does not bear out the proposition that negative political campaigns “work” in shifting votes toward those who wage them.

Overall, then, social science research provides some evidence that the mechanisms through which negative campaigning is supposed to work do in fact operate, but there is an overriding lack of evidence that negative campaigning itself works as it is supposed to. Intriguingly, the conclusion that negative campaigning is no more effective than positive campaigning holds even though negative campaigns appear to be somewhat more memorable and to generate somewhat greater campaign-relevant knowledge.

That negative campaigning can also rile up the supporters of the opposing side segues perfectly into the next reason why conservatives have been misled about the effectiveness of political advertising:

3. Claims made by Republicans in advertising are frequently derided and declared to be “lies” by left-wing journalists who have anointed themselves to the public as objective “fact checkers.”

A study conducted by the Center for Media and Public Affairs found that PolitiFact, a frequently cited website by journalists and campaigns, was more than twice as likely to rate claims made by the Romney campaign to be “false” compared to claims made by the Obama campaign. A separate analysis by the University of Minnesota found that PolitiFact was three times more likely to describe statements made by Republicans as “false.”

The bias of such “fact checking” organizations should come as no surprise given that they are primarily run by the same large media organizations which have pervasive ideological diversity problems. PolitiFact is owned by the Tampa Bay Times and many newspapers and television news organizations have created their own imitation operations.

Instead of always being about objective reality, however, sometimes the “facts” these websites claim to be providing are simply pronouncements on others’ opinions.

Weekly Standard writer Mark Hemingway, a long-time watchdog of the watchdogs, has provided several great examples of how biased and subjective the “fact checking” genre has become. The two examples cited below were written following the November 12, 2011 Republican presidential debate:

On Iran, former Massachusetts governor Mitt Romney suggested that the U.S. government should make it “very clear that the United States of America is willing, in the final analysis, if necessary, to take military action to keep Iran from having a nuclear weapon.”

Little did Romney realize that the AP is the final arbiter of America’s tactical military capabilities and can say with certainty that a military attack on Iran’s nuclear program should not be attempted. “The U.S. certainly has military force readily at hand to destroy Iran’s known nuclear development sites in short order. This is highly unlikely, however, because of the strategic calculation that an attack would be counterproductive and ultimately ineffective, spawning retaliation against U.S. allies and forces in the region, and merely delaying eventual nuclear weapons development.”

Also fortunate for the savvy news consumer, the AP apparently has a better grasp of what America’s intelligence agencies do and do not know than Newt Gingrich, a man who used to be third in line for the presidency and has received countless classified intelligence briefings.

At the debate, Gingrich suggested that there was room for improvement at America’s intelligence agencies, and noted in particular that we don’t have a reliable intelligence operation in Pakistan. The AP sprang to the defense of the CIA:

“The U.S. killing of a succession of al Qaeda figures in Pakistan, none more prized by America than Osama bin Laden, demonstrates that the United States indeed gets vital and reliable intelligence out of Pakistan. While it may have been true when Gingrich left government in 1999 that the CIA’s spy network was limited, since 2001 the agency has dramatically expanded its on-the-ground operations worldwide,” the AP “fact check” concluded.

The fact that bin Laden, the most wanted man on the planet, was living in a compound in Pakistan possibly for years may seem like a sign that our intelligence sources in the country leave something to be desired—but guess again, Newt.

The same negative media dynamic against Right-leaning messages is in effect outside of campaigns as well. In July of 2013, the conservative group Americans for Prosperity, a non-profit closely aligned with activist billionaires Charles and David Koch, announced it was going to spend $1 million to air a television ad across Ohio and Virginia criticizing the Obamacare health law. The ad was professionally done and could have been convincing to some people—if only it had not been denounced by legions of Left-leaning media types who took great exception to it.

Before we get into the criticism, first some background. The ad featured a woman identified as “Julie, mother of two.” She tells the story of her own experiences with the American medical system and how she is concerned about how the massive Obamacare law’s massive changes might affect her. Here is the full text:

Two years ago, my son Caleb began having seizures. The medical care he received meant the world to me. Now, I’m paying more attention. And I have some questions about Obamacare. If we can’t pick our own doctor, how do I know my family is going to get the care they need? And what am I getting in exchange for higher premiums and a smaller paycheck? Can I really trust the folks in Washington with my family’s health care? I think we all deserve some answers.

Whether or not Julie is an actor or a real person, the concerns expressed in the ad are valid. It is true that once the law went into full effect, Obamacare did prompt at least some employers to stop paying for their employees’ medical insurance. As left-leaning writer Sarah Kliff admitted in 2011 it would make some sense for employers since “dropping coverage is a pretty decent deal: A company would see its health care costs reduced by over 40 percent.”

As of today, larger companies do not seem to have cut back much on employee health insurance subsidies but smaller businesses have. According to the Kaiser Family Foundation, 68 percent of them offered insurance in 2010 while 56 percent did in 2015. All of that could change with economic conditions, however, as the New York Times noted in April. A downturn could mean significant cutbacks in employer generosity.

It’s also true that people whose employers have cut back on health coverage would be unlikely to be able to continue seeing their current doctors.

Additionally, it is also true that health insurance premiums for everyone except low-income people went up because of Obamacare’s mandating of numerous extra services in private plans and also its requirement that insurance companies cannot deny anyone from purchasing coverage. People who cannot afford to pay the rates of their current insurance providers are also unlikely to be able to remain with their current doctors.

While it’s true that not everyone will have to pay more for health insurance or be forced away from their preferred medical service providers, these were valid questions to raise at the time.

Conservative health care ideas (to the extent they even exist) have their own problems but it was legitimate for AFP to at least get people thinking about the current law. Left-leaning journalists strongly disagreed. Almost instantly after the ad campaign was announced, establishment media journalists set out to attack it.

New York Times editorial board member David Firestone accused AFP of “outright lying” in the advertisement, calling it “the worst kind of misinformation.” Christopher Flavelle of Bloomberg blasted the ad as “false,” “devious,” and “cunningly deceptive.” MSNBC denounced it as “misleading attack ad.” The Daily Beast condemned it as “nonsense” and lamented that the ad’s “emoting” could be persuasive to some people.

Taking a less histrionic approach, the Cleveland Plain Dealer dissected the ad in minute detail and rebutted it on several points while uncritically presenting the text of a rival ad from a Democratic group without any criticism at all. CBS News did the exact same thing.

Sarah Kliff, then of the Washington Post, who two years earlier had said that it was a “pretty decent deal” for companies to stop providing health coverage, changed her mind, asserting that “health insurance coverage under the Affordable Care Act will work pretty much like it does right now.” She now accused AFP of raising “a bit of a false hypothetical.” The next day, the Post’s self-styled “fact checker” Glenn Kessler gave the spot “Two Pinocchios,” claiming that “the AFP ad sets up a straw man.”

While there was some neutral coverage of the AFP advertising buy, its effectiveness was clearly decreased greatly by the copious amount of negative coverage it received from the press.

4. Money spent on ads is very short-lived. Even during a campaign, advertising spending has almost no lasting impact.

While campaigns have for decades run polls trying to track the effectiveness of their advertising right after they’ve engaged in it, what many non-academic political observers are not aware of is that there is an entire discipline within political science called “media effects” which has been looking at the impact that advertising has beyond just an overnight poll. What the scholars have found is completely the opposite from the conventional wisdom.

Perhaps one of the most widely believed notions in politics is that people actually remember campaign ads. “Early” ad spending has been credited by many political observers as what propelled the Obama 2012 campaign to victory against Mitt Romney. Bill Clinton’s former adviser Dick Morris believed it helped him overcome negative public opinion.

Actual scientific research of the longevity of advertising messages completely contradicts these claims. To the extent that ads have an effect, it lasts just a bit longer than the lifespan of a mayfly.

For several decades, the consensus within the academic world has been that campaign ads become almost completely ineffective after less than a week of airing. In 2006, this theory was put to the test by a group of social scientists with the cooperation of Texas governor Rick Perry and his chief political strategist Dave Carney who allowed them to use sophisticated research techniques to isolate advertising from the many other variables that can impact a campaign. The circumstances were ideal for study because at the time the analysis was conducted, there were no rival campaigns or organizations running ads.

The results confirmed what the political scientists had suspected: contrary to the conventional wisdom, neither radio nor television campaign advertising had “any detectable effect after two days.”

As George Washington University professor John Sides wrote in the New York Times, contrary to what most campaign consultants tell their clients and donors, they are as not nearly as sophisticated as the baseball database geeks first profiled in the book Moneyball:

Campaigns may be wasting millions of dollars running ads weeks if not months before election day, only to have any effects of those ads dissipate. Case in point: the approximately $20 million that Bill Clinton spent in advertising between July 1995 and January 1996 — months before the 1996 election. The mastermind of this strategy, Dick Morris, wrote that “the key to Clinton’s victory was his early advertising.” But there is little evidence that the ads mattered at all.

In other words, much of what goes into modern campaign advertising may be futile. Will this “rapid decay” convince candidates to husband their resources and unleash a barrage of ads on Oct. 30, 2012? Probably not. Might early advertising be useful for raising money or generating media coverage or other things besides moving poll numbers? Possibly yes, although here again there is only a little evidence, if that. Nevertheless, the fact a few eggheads have so spectacularly called into question the months-long television advertising campaign suggests how little may underlie the collected wisdom of the political cognoscenti.

Campaigns are spending a lot of money, but they are not playing Moneyball.

To the extent that campaign ads can be effective at all, it is for an extremely short amount of time. By contrast, operating a mainstream news outlet like a local television station, newspaper or magazine can actually be profitable and therefore self-sustaining.

More than three billion dollars was spent in 2012 on televised advertisements at the presidential level. The portion spent by opponents of President Obama has earned literally zero return on investment.

5. Every dollar in ad buys is a dollar into the hands of the very same media that attack conservative beliefs and candidates instead of being put to use building things that can last.

Even if the TV ads that conservatives love so much were actually useful at the presidential level, it seems like none of the organizations paying for them have ever stopped to consider just who they’re buying the ads from: the very same left-leaning media outlets that are constantly condemning them.

Republicans have, quite literally, been subsidizing the media arm of the political movement that opposes them.

This tactic would be considered madness if it were applied to any other form of human activity: Coke does not purchase advertising on Pepsi bottles, Harvard does not give endowment grants to Yale, and Ford does not seek to help defray the costs of Toyota customers. Anyone touting the advantages of such ideas would immediately be dismissed as insane and rightly so.

Of course some amount of advertising is needed but to insist that billions of dollars be spent in such a fruitless manner to the exclusion of any other idea is absurd.

Just as no amount of Pepsi advertising dollars will ever be able to convince Coke drinkers that they should switch beverages, Right-leaning philanthropists need to realize that no amount of money spent on campaign ads is somehow going to make Americans become more conservative.

Over-relying on advertising is also a less effective way of developing conservative talent. News websites or TV stations need a lot of writers, editors, and technicians in order to function. A political ad rarely needs more than 20. Besides having a much lower return on investment, ad buys cost far more than media properties, particularly online ones.

In 2012, the Mitt Romney campaign and its associated political action committee spent nearly $1 billion trying to defeat President Obama. Outside groups spent several hundreds of millions, if not over a billion more dollars on top of that.

If even one thousandth of the approximately $2 billion that was spent had been invested to create or boost center-right media outlets, the net effect would have been far more significant per dollar. The odds are pretty good that the operations funded would still be in existence today. They might even be profitable like the Daily Caller is.

Conservatives love to complain about the liberal leanings of the mainstream news and entertainment companies but if they were actually serious about their gripes, they would stop funding them and do more than just complain. The only way to beat the media is to become the media.

You May Also Like


  1. 1

    Simply fantastic. Where is this kind of research on the conservative websites?

    Actual political science? Ye gods!

  2. 2

    First of all, it’s nice to see something by Matt Sheffield publishing something again. It’s not that he hasn’t been publishing, its that we’ve been traveling in different circles for a while. In regards to campaign advertising, my gut is telling me that, if Donald Trump wins this, and doesn’t spend $1 billion to do it, the whole landscape of political advertising will be changed forever. I’m not sure if he can pull this off, but I do see a lot of opponents spending millions against him and he not spending much, and winning. Hillary has spend over $50 million against him and his poll numbers are rising.

    It’s going to be something beautiful to watch. Anyone want to bet the Democrats won’t be going after Citizen’s United after this but just might go after Facebook and Linkdin? I’m not sure where this all goes, but I do know that things are changing.

  3. 3

    The consultants control the GOP. The donors on both sides are nothing but addled old people who do what their minders tell them to.

    None of them has any idea what to do with the money. I hope some of them read this article and call up Mr. Sheffield.

  4. 4

    This is a very good article but it doesn’t quite get into where ads might be good. It says people have to do them but doesn’t give recommendations as to why. It’s really long though so I guess maybe that’s why. Just a suggestion.

  5. 5

    Not only should conservatives create and support friendly media outlets, they need to be involved in TV and movies. Propagandized, statements from “global warming,” to evil CEOs, to distortingting Christian beliefs, is dropped into presentations in such a matter-of-fact manner that viewers do not notice the subtle impact – but over time it adds up and influences.

Comments are closed.