Summary of Ad Effects, Essay Example

Advertising has always had a negative or positive effect on viewers. The very development of television ads were to use both stimulate viewer’s perceptions and influence consumers to purchase their products. According to research consumers are exposed to over three thousand television advertisements every day. It is no wonder that the political pundits have taken to increase advertisements in the media in order to influence and inform potential voters. “Presidential elections provide both an important context in which to study advertising and a setting that mitigates the challenges of dynamics and endogeneity.” (Gordon, Hartmann, 2013) In the study conducted by Brett Gordon and Wesley Hartmann, they sought out to test the lengths at which advertising effected the public during presidential elections. This paper is developed to summarize, “Adverting Effects in Presidential Elections”, in order to provide the purpose, the study, and the future research.

Advertising has long thought to be influential, firms and companies spend billions of dollars a year in order to capture the public’s attention for a mere 30-seconds. When studying advertising focus points to comprehending advertising effects and why it works. “Economics and marketers debate whether advertising is informative or persuasive: marketers assess its effects on intermediate measures such as brand recall: and political scientists wonder if negative advertisement depress voter turnout.” (Hartmann, Gordon 19) The quest for this information is still elusive to research and the studies conducted are still not concrete. Studying advertising, one of the most essential variables is discovering its influential power of political elections. Each year the political ads from political pundits increases each year, and experts alike wonder if the power of advertising contributes to the choices of voters, reasons for campaign spending, and fundraising. “An election aggregates the decisions of many into a single outcome with far-reaching consequences.” (Gordon, Hartmann 19) The study conducted by Gordon and Hartmann (2013) analyzed the general elections of 2000 and 2004, that effected the market-level advertising on country-level vote shares. (Gordon, Hartman 20) Within their study to measure cleanly the possible effects they have included an extensive set of fixed effects at the market-party level. (Gordon, Hartmann 20)

The research question posed by the duo is, the advertising effectiveness in presidential elections. The two challenges they had with the research and research question in general was estimating the effects of econometric endogeneity and the convoluted effects of the past and presented advertising. The endogenetiy problem is spurned by the correlation between unobservables and the casual effect in isolating the problem. In order to address the issues they took advantage of elections not being held during odd years to use the prior year’s prices in advertising as cost instruments. The second challenge was solved by focusing on major metropolitan areas that have higher advertising prices, and the fixed effects used showed the shift inference on how within market changes the prices of advertising levels and vote shares. (Gordon, Hartman 20) The study used pooled candidate-shared observations from two elections across the country, observing 9,576 advertising exposures and vote shares. The hypothesis of how these advertising positively affected voters, was estimated to show, “a robust positive advertising effect across a number of specifications, including numerous exogenous control variables and their interactions with political party dummies.” (Gordon, Hartmann 20)

The study composed of data measured at a state and country level in order to observe all media-market levels. The data comprised of 78 percent of the country’s population in regards to advertisement placement. Each ad was observed by the dates and times in which it aired, the length in seconds, the supported candidate, and the sponsoring group. The total spend on advertising in 2000 and 2004 was, $168 million and $564, respectively. They expressed their GRPs, gross rating points for key advertising variables that measures the number of exposures per capita. In subsequently measuring the endogenenty of the candidate advertising, they considered the decision process of to choose ads that were instruments in helping them choose suitable markets. They study observe the significant variations in which they choose to advertise during the day, and the prices across the elections. In measuring voting, they observed the data from 1,594 counties, the size of the voting age population, and all possible candidates. The additional control variables uses include; local political preferences, voter turnout but not candidate choice, demographic and economic variables, and candidate-specific local variables. The group used a static aggregate discrete choice model of demand for the political candidates in order to reflect the utility of the voters for each candidate, and the mechanism precisely used in which advertising affects the candidate choice. (Gordon, Hartmann 27)

The results of the paper showed that advertising had robust positive effects in the case of general elections for presidency. The instrument variables, the fixed effects, and controls observed within the study impacted the estimated advertising coefficient. Their findings indicated that adverting is influential in shifting voter’s perspective in terms of electoral votes in multiple states. Their research contributes to measuring political advertising by addressing the endogeneity problem of advertising, and the model used combines the voter’s decision, and the decision of candidate choice. In terms of future research, they did not consider the positive or negative effects of advertising, and did not model the voters’ expectations about the potential outcomes of the elections, and how this would alter voting decisions.

Works Cited

Gordon, Brett R., Hartmann, Wesley R. Advertising Effects in Presidential Elections. Marketing Science. Vol 32. No1. Pg 19-35. 2013. Print.