Does Politics Still Stop at the Water’s Edge?
Senator Arthur Vandenberg (arm raised) during the 1948 GOP Convention
Source: Thomas D. McAvoy/The LIFE Picture Collection/Getty Images, https://www.npr.org/sections/parallels/2015/03/10/392095906/taking-u-s-politics-beyond-the-waters-edge
Senator Arthur Vandenberg (R-MI) once quipped that “politics stops at the water’s edge.” The point was that Americans fight like cats and dogs over domestic issues (e.g., unemployment and taxes), but set aside their differences when it comes to foreign policy (e.g., war and foreign trade). Vandenberg, himself, was a case in point. Despite the senator’s strong opposition to the New Deal, he supported the Truman Doctrine that defined American foreign policy for the next 40 years.
Does politics now extend beyond America’s borders? Has the polarization that has characterized American domestic politics for the past 20 years now seeped into foreign policy? Answering these questions is tougher than one might think.
This essay will explore two traditional ways of looking at polarization before providing our own “benchmarks" method.
THE LONGITUDINAL METHOD
The traditional way to see if the American public has polarized is to look at the same survey question asked at different points in time. Here is one example from the American National Election Study (ANES)…
There are four interesting things about this chart.
The greatest period of polarization between Democrats and Republicans occurred during the Reagan-George H.W. Bush years.
Most Americans reject an isolationist foreign policy (the total percentage of Americans who prefer isolationism never gets above 30 percent from 1956-2020).
Democrats have historically been more isolationist than Republicans, but that is no longer the case.
Republicans have grown more isolationist since 2004.
Is the question about “staying home” a good indicator of Americans’ foreign policy preferences? For a couple of reasons, the answer is “probably not.” First, the question is pretty vague. Does “staying home” mean not fighting foreign wars? Not engaging in international trade? Not providing foreign aid? Second, the questions that were important back in the 1950s (e.g., the Soviet Union and containing communism) are not the same ones that are important today (e.g., what to do about climate change and refugees?).
THE SOURCE CUE MODEL
A second way to assess polarization is to examine the effects of partisan source cues. Source cues are pieces of information that help us answer complicated questions. In an era of hyperpolarization, inserting a president’s name into a question provides a powerful source cue. However, presidents’ names were not great source cues in eras when Americans were not so polarized.
Consider source cues in Americans’ response to how Bush and Obama respectively handled the situation in Afghanistan:
Republicans loved the job Bush was doing in 2007 (77% approve), but hated the job Obama did in 2009 (17% approve). Conversely, Democrats disapproved of Bush’s handling of Afghanistan in 2007 (13%) but gave Obama high marks in 2009 (65%).
There are two explanations for the ol’ partisan fliparoo. First, Obama might have changed U.S. policy in Afghanistan in ways Democrats liked but Republicans didn’t. This seems unlikely since the biggest change in U.S. policy did not occur until 2010. A second more plausible explanation is that Americans are reacting to the occupant of the Oval Office, not their policies. In other words, “Bush” and “Obama” provides partisans with all the information they need to answer a question about Afghanistan. (It is also instructive that Independents—those without a strong partisan leaning—remained relatively unchanged: 34% approved of Bush; 33% approved of Obama.)
Estimating the degree of polarization from questions like this is great, but, unfortunately, they don't come along very often and can be like comparing apples to oranges.
THE BENCHMARKS METHOD
Along with my co-author (Brianna Newport), we propose a new way to estimate polarization in American foreign policy. We call this the “benchmark method.” Here’s how it works:
We gathered all relevant surveys in 48 of the most critical foreign policy cases over the last 50 years, a process that cumulated in a dataset of over 28,000 questions.
We classified each case as a foreign policy crisis or a non-crisis.
We then looked at the level of public agreement for each survey question, with opinions coded as “unified” (60 percent agreement and above) or “divided.” The more “divided” questions, the more polarized the American public.
Finally, we sought to discover whether some stages of foreign policy were more polarizing than others. To do this, we classified each question into four categories:
Knowledge and Attention
Problem Definition
Options
Decisions
The resulting dataset allows us to examine levels of polarization over time, the difference in public perceptions of crises and non-crises, and which stages of decision-making tend to be the most polarizing.
Here’s what we found.
First, Americans tend to be more unified on foreign policy crises (43%) than non-crises (37%), which is consistent with the rally-around-the-flag effect.
Second, the trend is toward less unity of opinion in foreign policy crises and non-crises (i.e., Americans are more polarized now than they were back in the day). Still, the trendlines are not as steep as we expected.
Third, the most polarizing crisis (Iraq War 2004-08) and non-crisis (reauthorization of NAFTA and the Trans-Pacific Partnership) are relatively recent events, which suggests foreign policy polarization is fairly new. Still, other 21st Century events (e.g., terrorism and Kyoto Protocol) proved rather unifying, which works against the polarization hypothesis.
Finally, the level of polarization differs by the stage of decision-making. In crises, for example, Americans are far more unified in defining a problem (50% unified) than in approving of a decision (40%). Our tendency to look only at decisions—usually the most polarizing aspect of any case—would then lead us to overestimate foreign policy polarization.
Conclusion
Our “benchmarks” method paints a more muddled picture of polarization in American foreign policy than we expected to find. While the trendlines show a decline in public unity over the years, the slopes look more like green runs on a ski hill than double-black diamonds. Moreover, if we examine those overlooked aspects of foreign policy—i.e., the agenda-setting, problem definition, and option-generation stages—we often find broad areas of agreement. Put differently, a myopic look at policy decisions overestimates the degree to which Americans are polarized.
Is the American public more polarized now than they were in the past? Yes. Is foreign policy polarization as pronounced as domestic policy polarization? No.
