While campaigning for the White House, U.S. President Joe Biden promised Americans that he would reenter the nuclear deal with Iran, officially known as the Joint Comprehensive Plan of Action (JCPOA), so long as Tehran returned to compliance with the agreement’s original terms. Recent polling indicates that this would be a popular move: a majority of the American (and even Iranian!) public would support a joint return to the JCPOA, suggesting that Biden has a solid popular mandate to renew the pact. Yet nearly five months into Biden’s presidency, and after several weeks of indirect talks in Vienna, however, reconstituting the agreement has proved elusive. Why have we not yet witnessed more significant progress toward a reconstituted JCPOA?

One major roadblock lies in how past behavior by both the United States and Iran hangs over the negotiations. Many Iranian elites do not trust Washington following its abrogation of the original deal under President Donald Trump (not to mention American support for the coup that overthrew Iran’s democratically-elected prime minister in 1953), while many American elites are wary of Tehran given its history of deceiving the IAEA, facilitating ballistic missile proliferation, and sponsoring terrorism. 

Given this context, influential observers on each side are skeptical that the other party can credibly commit to upholding its end of the bargain. The deadlock is consistent with conventional wisdom about the importance of a state’s past actions to how observers assess its reputation for keeping commitments, and in turn, measure the credibility of its current and future promises. As Thomas Schelling famously argued in Arms and Influence, events and issues in international politics are highly interconnected: a failure to meet a commitment has a redounding effect, not only on the issue in question, but also on a state’s other foreign policy endeavors, so states must honor their pledges if they expect others to regard them as reliable actors.

Yet as Robert Jervis pointed out in Dominoes and Bandwagons, the interconnectedness of events is not a point of fact, but rather a function of individuals’ beliefs about the nature of potential spillovers between issues and events. As a result, some observers will care more than others about whether another party to a potential agreement has kept its commitments in the past. The implications are particularly relevant to negotiations over the JCPOA, which has proved politically polarizing as a function of individuals’ prior beliefs— moderates in both the U.S. and Iran have tended to support the agreement while conservatives have decried it.

We examined how observers’ beliefs shape their policy views on this issue through a survey experiment fielded on a representative sample of the U.S. public and a recent snap poll conducted by the Teaching, Research, and International Policy (TRIP) project at William & Mary. The results allow us to draw illuminating comparisons between how mass audiences (proxied by our survey) and elites (proxied via the TRIP sample) think about past behavior and the interconnectedness of events in the context of the Iran deal. 

In the experiment, we asked about 3,000 respondents to evaluate one nation’s hypothetical commitment to lift sanctions on a rogue dictatorship in exchange for the latter’s commitment to dismantle its nascent nuclear program. We varied whether the country proposing the commitment was identified as the United States, a parliamentary democracy, or another dictatorship. We also deployed a novel scale to measure respondents’ beliefs about interconnectedness by capturing their level of agreement (on a five-point scale ranging from “strongly disagree” to “strongly agree”) with the following statements:

• In international affairs, states are more likely to cooperate with each other than to come into conflict.
• When one state acts aggressively, other states need to respond with force to avoid setting a bad precedent.
• When states cooperate on one issue or in one area, this creates a positive feedback loop that enables them to cooperate on other topics.
• A country’s defeat or retreat on one issue or in one area of the world will lead to new challenges from that country’s adversaries.

We sum all four items to create a composite index on which “high types” for interconnectedness fall in the top quartile of the scale’s distribution while “low types” fall in the bottom quartile. We consider high types as those who see strong interconnections between world events, while low types are those who do not see the world in this way. With the gracious assistance of TRIP’s faculty and staff supervisors, we also embedded this battery of questions within their April 2021 snap poll.

In both samples, respondents were asked comparable outcome questions about their support for the hypothetical or real agreement on behalf of the proposing state. Following the experiment, respondents indicated how strongly they supported or opposed a) the proposing state’s commitment and b) the proposing state following through on its pledge (provided that the rogue state upheld its end of the bargain). For the TRIP survey, respondents indicated whether they supported or opposed President Biden’s efforts to engage Iran in informal discussions about JCPOA compliance. 

We regressed these outcomes on respondents’ beliefs about interconnectedness, subsetting the experimental sample to just those who were told that the proposing country was the United States to maximize comparability with the TRIP data. We include partisan identification and gender as potentially relevant covariates, with the expectation that partisanship in particular should be an influential predictor given how Democrats and Republicans have split on their support for the Iran deal.^[1]While these results cannot be interpreted causally, they not only indicate a robust relationship between beliefs about interconnectedness and an important policy issue but also demonstrate that this relationship looks different among elites versus the public. 

The regression analysis indicates that high types in our experimental sample are significantly more likely than low types to support the proposing state’s commitment and to say that the proposing state should keep that commitment — the coefficients on our interconnectedness scale are positive and statistically significant (with p<0.05) regarding respondents’ support for the commitment and whether they think the proposing state should abide by its commitment. Yet high types in the TRIP sample are significantly less likely than low types to support indirect engagement with Iran over JCPOA compliance, with the coefficient on our interconnectedness scale entering the model as negative and statistically significant (also with p<0.05). Both sets of results are robust to controlling for gender and partisan identity, suggesting that our variable of interest is a better predictor of respondent’ attitudes than either their gender or whether they identify as a Democrat, Independent, or Republican.

We interpret these findings as illustrating significant differences between elite and mass audiences when it comes to U.S. policy toward Iran. We speculate that interconnectedness might mean different things to these audiences based on their own experiences. The results for the general public are consistent with other polling that demonstrates majority support for reengaging with the Islamic Republic. Perhaps this is because mass audiences are thinking about the future impact of adhering to the present commitment and expecting better relations between the U.S. and the hypothetical rogue state (which we do not identify in the experiment, but may have sounded like Iran to respondents) as a result. 

Meanwhile, the results for our elite sample are in line with a much more skeptical view of Iran’s likely future behavior. It may be the case that as close observers of international politics, members of the TRIP sample are more cynical about the chances for Iran’s future compliance given its historical behavior — perhaps they fear that Iran will dupe the United States into revamping the agreement while maintaining a clandestine weapons program. 

What is even more remarkable about elite resistance among high types is that it remains a significant predictor despite the sample comprising mostly Democrats, who we know otherwise tend to support the Iran deal (given that TRIP surveys academics, the underlying population skews heavily in this direction, without about 70 percent identifying as Democrats). To the extent that elite academics are a reasonable proxy for elite policymakers, these results could point to one reason that the Biden administration has not rushed to conclude a new agreement with Iran. 

Though recent research points to remarkable similarities between elites and the public on foreign policy questions, our findings suggest that significant differences in beliefs do persist between these populations. While more research is needed to explore the sources and consequences of these differences, for now we simply posit that beliefs about interconnectedness are an important but previously underexamined foreign policy disposition that may explain key differences in what elites and the public are seeking from the future of U.S. relations with Iran.

^[1] Interconnectedness is a 3-point scale with ‘Low’ coded as -1, ‘Medium’ coded as 0, and ‘High’ coded as 1. Party ID 3-point scale with ‘Democrat’ coded as -1, ‘Independent’, coded as 0, and ‘Republican’ coded as 1. Gender is a binary where ‘Male’ equals 1 and ‘Female’ equals 0.