![](https://webarchive.library.unt.edu/dentonfracking/20141029180001im_/http://static01.nyt.com/images/2014/10/29/upshot/29up-vavreck/29up-vavreck-master675.jpg)
Nearly everywhere in America tonight, people will turn on their televisions and grumble about all the political ads. Here’s why you should welcome those ads: Evidence suggests they’re important to our democratic health.
A functioning democracy needs an electorate that makes informed choices. Much as we dislike them, political ads, especially in midterm elections, convey information to voters about candidates, particularly those who are unknown to most people.
For example, evidence from recent midterm elections showed that in places where candidates advertised with greater frequency, voters on average knew more objective things about the candidate. The effects are notable for something as straightforward as helping voters identify who is actually running in the race. And just like campaign spending generally, challengers’ ads have greater impact than those of incumbents.
Here is an example of how this works: After the midterm elections in 2006, the political scientists Seth Hill, James Lo, John Zaller and I showed 3,002 people pictures of their sitting member of Congress and the person running on the other side. In open-seat races, we showed both major-party candidates. We used representative samples from nine media markets in the Midwest covering 127 candidates in 65 districts. As a measure of political information, we asked people to tell us the names of the people in the photos. We gave people no choices — only a blank line to fill in whatever came to mind. We included some other politicians and notable people, too.
Some people were able to name the politicians in the photo. Others knew a little about the person, but couldn’t come up with a name. They wrote things like “that lady who ran for Congress.” A few offered comments indicating they recognized the person in the photo and had an opinion about that person’s politics. We counted all such responses, which indicated some degree of familiarity with the politician in the photo, as “correct” for our purposes.
Immediately after the midterm election, 46 percent of the people in the sample in races where there was an incumbent seeking re-election could correctly identify their sitting member of Congress from just a photograph, but only 21 percent could recognize the person running against the incumbent. That 25-point gap is why challengers have more work to do in their campaigns. It’s also one of the reasons their dollars go further than incumbents’ dollars do.
Because people did not do as well recognizing candidates in the open-seat races, the overall share of people who failed to identify any candidate was 57 percent. On average, 23 percent recognized just one of the candidates in their district, and 20 percent recognized both.
Midterm Elections 2014
![](https://webarchive.library.unt.edu/dentonfracking/20141029180001im_/http://graphics8.nytimes.com/images/2014/09/08/us/midterms-promo-image/midterms-promo-image-thumbWide.jpg)
The latest news, analysis and election results for the 2014 midterm campaign.
But there is considerable variation in the number of people who can recognize their incumbent member of Congress across districts, from lows of zero in places in Michigan to highs of 68 percent in neighboring Ohio. . These differences may be easy to understand for incumbents — some members go home more often or have been in office longer. Even challengers may be more or less familiar depending on their previous experience in the district.
In fact, for challengers, more of the variation in recognition took place across media markets than within them. That means that factors like respondents’ levels of education or income, race, gender and age, which we would expect to explain a good deal of the variation in whether someone can identify a political candidate or sitting member of Congress, explained less of the pattern on recognition of the photos than where people lived.
Media market designations — and the level of political advertising within those markets — affect whether people can identify who the candidates are for office more than typical characteristic about the voters themselves. The effects of campaign ads, for example, can trump the average effects of education in the electorate.
Put differently, in terms of how much information people have about the candidates running for Congress, it matters less whether a person has a college degree than whether he or she lives in a media market with lots of political advertising.
People learn basic information about elections from ads.
The level of challenger advertising in a media market explains nearly 80 percent of the variation in recognition of the candidates across the markets, controlling for demographics like age, race, gender, income and employment status. For every additional 10 ad viewings per capita in a media market, recognition of the challenger increased by 20 points on average, to 39 percent from 19 percent.
For incumbents, the payoff for the same 10 ad viewings was much smaller: only a six-point increase, to 58 percent from 52. Moreover, things like the number of news outlets in the market or corporate cross-ownership of media platforms failed to explain any of the patterns of knowledge. It’s not the structure of news in the market that was informing people, it was the ads.
As was the case with the patterns of campaign spending we showed earlier at The Upshot, one of the things people hate about the process is actually one of the ways challengers can level the playing field.
Don’t hate the players, hate the game.
The Upshot
In a weekly e-mail newsletter, get the best of The Upshot's analysis about politics, policy and everyday life.