The surprising finding that "cultural worldviews" don't explain people's views on gun control.

• Author: Fremling, Gertrud M.
• Position: Response to article by Dan M. Kahan and Donald Braman in this issue, p. 1291

Using the General Social Survey, Kahan and Braman show that "cultural worldviews" only explain about 1.6 percentage points of people's views on gun control. Assuming that their indexes properly measure these "cultural worldviews," the results are very striking and surprising. In addition, 92% of the variation in people's views isn't even explained when all the factors that they account for are included. While they don't test the hypothesis, it is very conceivable that some of the 92% can be explained by what information people receive. If so, the returns to research could be quite large.

INTRODUCTION

Do people's values and political perspectives influence their positions on gun control? Certainly, there is no doubt about it. But facts must also matter. To say that only people's values matter seems oddly circular, because it fails to explain the origin of these values. If values do not come from facts that we encounter--either through our own experiences, those we learn from other individuals, or from more evidence of a scientific nature--where would they come from? Are values the result of pure randomness? Not likely. There must be some contributing factors. Are values the result of innate human instinct? Possibly some. But if one were to argue that attitudes originate from religion, politics, or moral philosophy, would not these attitudes ultimately originate from the perceptions held by prophets, politicians, and philosophers? And these perceptions would not be based on reason alone, but on these individuals' experiences as well as on experiences and facts related to them by other individuals.

While Kahan and Braman's article focuses on gun control, their claim that empirical work is of little value (1) must surely apply to other issues. Yet, the importance of facts in altering views seems undeniable. Surely the failure of communism and its fall in the former Soviet Union and Eastern Europe affected views on government ownership of production. Today few economists deny that price controls cause shortages. (2) The vast majority also believe that increasing the money supply is the primary reason for inflation. (3) The consensus is so overwhelming today that it is hard to remember how contested these ideas were just four or five decades ago. (4)

Scientific discoveries reflect another form of empirical work that has obviously revolutionized our thinking about the universe and the place and role humans occupy. Science has generated much bigger upheavals than mere "attitudes" about social issues. For example, as evidence gathered that the Earth circled the sun, we could no longer view ourselves as being the center of the universe. Likewise, the concept of evolution and the evidence behind it profoundly changed our views on human nature itself.

Kahan and Braman should be lauded for attempting to provide empirical evidence using data obtained from the General Social Survey (GSS). Data gathering and statistical analysis should be done more, not less, frequently. Of course, that is our perspective. The irony is that they try to use empirical evidence to show that empirical evidence doesn't matter.

If found correct, their bold hypothesis would surely shake up academia and political debate. Philosophy and religion, which are more heavily based on reasoning and ethics, would increase in prominence, while economics and, in particular, statistics would become less relevant. Unfortunately for them, their results do not support their claims. Indeed, as we demonstrate, Kahan and Braman's results show the very opposite of what they hope.

1. THEIR EVIDENCE

   Kahan and Braman reach two closely related conclusions: (1) "individuals' attitudes toward gun control are derivative of the type of social order they prize"; (5) and (2) "those interested in resolving the gun debate should turn their attention away from quantifying the consequences of gun control." (6) These conclusions are based on the two regressions reported in Table 1 of their article. (7)

   ...