ACSI Presidential Election Survey: A Measure of Voter Satisfaction With the Major Party Candidates and Their Campaigns
Week of September 5 to September 9, 2016
The American Customer Satisfaction Index (ACSI) launches a new study tracking “voter satisfaction” with the race for U.S. president, to be updated each week with new data until November 7, 2016 (the day before the election).
Customer satisfaction is a proven predictor of purchases of goods and services; therefore, it should have predictive power for presidential elections as well. Like consumers, voters may pick candidates based on the “expected satisfaction” that a candidate will deliver. Expected satisfaction, in turn, is predicted by current satisfaction. Accordingly, the ACSI will track expected and current satisfaction with the two major party candidates, Hillary Clinton and Donald Trump.
In markets for goods and services, current satisfaction is determined by a customer’s “consumption experience.” In the absence of any such actual experience prior to the inauguration of a new president and a period of actual governing, the ACSI uses a proxy that measures satisfaction with each candidate’s campaign. From these two measures—expectations and satisfaction with the candidates and their campaigns—both the breadth (“market share” or voter share) and depth (satisfaction) of support are derived for each candidate. An algorithm for estimating voter share and satisfaction is used to group registered voters into “strong” and “weak” supporters of each candidate and to identify both the size and satisfaction levels of those groups.
Beginning with data from the week of August 1 to 5 for the first wave and ending with the week of September 5 to 9, a total of 5,311 interviews were conducted thus far. During the first week of August, voter satisfaction was 74 for Clinton and 72 for Trump. These results do not represent percentages but rather ACSI scores on a scale from 0 to 100. The corresponding estimated voter share (or market share) in week one for Clinton was 49% versus 39% for Trump. Thus, not only was Clinton’s satisfaction score higher than Trump’s, her share of the vote also was substantially higher.
Following the first week of August, voter satisfaction for Trump slipped, rebounded, and peaked, and then declined for two consecutive weeks. During the same period, Clinton’s voter satisfaction was mostly flat, but then also declined for two weeks. According to the most recent data, Clinton’s satisfaction score has dropped to 73 (-1), while her voter share has receded 2%. Meanwhile, Trump’s satisfaction is unchanged since the first week of August, although he has gained in voter share (+3% to 42%). Nevertheless, Clinton continues to lead in both voter satisfaction (73 vs. 72) and voter share (47% vs. 42%).
Unlike traditional trial heat polls, measuring current and expected satisfaction with the major party candidates allows researchers to group voters into “strong” and “weak” support groups. In the diagram, “strong” supporters are those defined as being much more satisfied with one candidate and much less so with the other. “Weak” supporters are defined as those only somewhat more satisfied with one candidate over the other. “Undecideds” express equal satisfaction/dissatisfaction with the two candidates.
For Clinton, 35% are strong supporters. For Trump, 30% are strong supporters. When combining strong and weak supporters, Clinton currently has a voter share of 47% and Trump has 42%. 11% are equally satisfied or dissatisfied with the candidates.
The following table shows how support for each candidate has fluctuated over the past six weeks. For Trump, the share of strong supporters has risen from 27% to 30%, while Clinton’s has actually declined (to 35% in week six from 38% in week one). Undecideds have shrunk from 13% of the total to 11%, although the number of undecideds is actually higher now than in the prior week.
The following tables depict voter share for each candidate for the week of September 5 to September 9 among various demographic groups. Clinton holds a strong lead over Trump among respondents 18 to 34 and a slight lead among voters 35 to 54. The two candidates are nearly even among voters 55 and above. Similarly, Clinton holds a big lead among women, while enjoying a smaller lead among men.
Clinton holds a massive lead over Trump among African Americans and a strong lead among Hispanics, but among white voters the two are tied. Finally, respondents with less than a college degree break decisively for Trump, while those with a college degree or more favor Clinton by almost 20 percentage points.