Taylor Rawson and Dr. Darren Hawkins, Political Science

Introduction

Scholars have long researched public support for foreign aid within donor countries. Democratic governments are held accountable by their constituencies, and as a result, public opinion is of importance. However, although the United States is by far the largest aid donor, the American public grossly overestimates the size of the U.S. foreign aid budget and is largely skeptical of foreign aid in general. In a preliminary experiment, conducted in late 2014, Dr. Darren Hawkins and I explored how citizens react when they are given certain kinds of information and/or opinions regarding foreign aid. This experiment explored the different foreign aid arguments that we identified as existing in public discourse. We included arguments both for and against foreign aid and presented these arguments as a series of five facts, a one-statement opinion, or both facts and an opinion. We found that facts and fact/opinion treatments elicit stronger reactions than pure opinion treatments. Additionally, we found that the treatments that generated the most support for aid were those that emphasized the relatively low cost of aid and the overwhelming need for aid, while the treatments that most effectively decreased support for aid were those that emphasized the relatively high cost of aid and the high levels of corruption associated with aid projects.

My ORCA project allowed us to conduct an additional, two-part experiment to refine and extend these original results. The first part of this experiment allowed us to again examine the relationship between opinion and fact & opinion frames, and our second part allowed us to test both positive and negative fact & opinion frames against each other.

Methodology

I, along with Dr. Darren Hawkins, designed a survey experiment using the Qualtrics platform and then administered it in June of 2015 to a representative sample purchased from Qualtrics. In this experiment, subjects were presented with one of twelve treatments or a control condition. The treatments were each variations of the four arguments that we found to be most influential in moving support for foreign aid in our first experiment. These four arguments—two in favor of foreign aid and two against—include the ideas that foreign aid is inexpensive, there is need for foreign aid, foreign aid is lost to corruption, and foreign aid is expensive. Our first four treatments isolated each of these arguments in treatments containing a series of five facts paired with an opinion frame, while the second four treatments presented the same treatments as an opinion frame only. Here, we sought to replicate the finding that opinion frames, when paired with facts, are far more powerful in moving public opinion than opinion frames alone—which goes against much of the existing literature regarding opinion frames and their effects.

The final four treatments of this experiment paired a positive argument with a negative argument and presented both as a series of five facts and an opinion frame. Here, we sought to determine whether positive or negative information is more powerful. These treatments paired each of the positive arguments with the negative arguments, resulting in an additional four treatments. Our control condition, in contrast, contained no information. All respondents were asked a series of demographic questions and, following the condition, their opinion of whether U.S. foreign aid spending is too high, too low, or about right.

Results

First, we confirmed the results of our original experiment and again found that treatments that include facts are far more powerful in moving public opinion than are opinion frames alone. In fact, all four of our treatments presented as a series of facts in addition to an opinion frame were statistically significant at the 95% level, while those presented as an opinion statement alone fell short of statistical significance. We confirmed those results using both t-tests and logistic regressions.

With regard to the second portion of this experiment, we found that pairing both positive and negative sets of information regarding foreign aid spending and presenting both to respondents did not result in significant movement in public support for foreign aid. As a result, pairing arguments both in favor of foreign aid and against it negates the effect either argument alone could have had on public opinion regarding foreign aid.

Discussion

Perhaps the most intriguing finding of this experiment is that opinion frames alone do very little to alter public opinion, at least with regard to foreign aid spending in the United States. There has been much research regarding the effects of framing, and this finding challenges that body of literature.

It is similarly intriguing that, of the four treatments I employed in my ORCA project, three were financial in nature. Again, these four treatments were chosen because of their high significance levels in our first experiment, which perhaps reflects a popular misunderstanding of the nature of foreign aid spending in the United States. American citizens seem particularly pliable when presented with information regarding the cost of foreign aid—whether high or low—or its loss at the hands of corrupt foreign leaders. As a result, I posit that more effort on the part of the government to inform its citizenry of the true nature of foreign aid spending could improve its perception drastically. Similarly, arguments backed by facts seem the most powerful in altering support, and the government and/or special interest groups would be wise to employ them in their advertising campaigns and other public appeals.

Conclusion

In conclusion, Dr. Darren Hawkins and I studied the effects of particular arguments, and the way that they are presented, on the foreign aid debate. We employed the four most influential arguments that we found in a previous experiment and compared the effects of each when presented as a series of facts and an opinion frame or solely an opinion frame. In addition, we paired each of the two arguments in support of foreign aid with each of the arguments against foreign aid in order to explore whether positive or negative arguments are more powerful. We found that arguments presented as an opinion frame backed by facts are significantly influential, while arguments presented as an opinion frame alone are generally not statistically significant at all. Secondly, the effect of either positive or negative information alone is negated when the arguments are paired together. As discussed, these findings have implications for both the existing political science literature and also public policy.