It is important to read beneath the headlines and initial interpretations of a new NBC News/Marist poll that surveys public opinion on President Obama’s proposal for a congressional authorization to fight ISIL. The survey asked the following question:

“President Obama has asked Congress to vote to allow the use of U.S. military force against ISIS, the Islamic State group. The president is requesting authorization for three years, with no geographic limitations. The president is also asking for flexibility for limited ground operations by the U.S. military, but rules out deploying a long-standing ground force. Do you want your member of Congress to vote for or against this authorization?”

The current coverage focuses on two findings:

1. A majority of American said they support the President’s proposal: Americans (54%) favor the President’s proposal and a minority (32%) are against it (13% are unsure).

2. A potentially surprising discrepancy between Republicans and Democrats: Only 52% of Republicans said they support the authorization request while 60% of Democrats support it.

Let’s tackle the second one first. Greg Sargent over at the Plum Line interprets it to mean that “Republican voters appear at odds with GOP lawmakers on this topic. The latter have been arguing that, if anything, Obama’s request is too limiting. … But a bare majority of Republicans supports the limits in the authorization Obama proposed, such as they are.”

I’m not so sure. Many of the Republican respondents may simply disfavor the President’s proposal because they believe it is too limiting. Indeed, that’s not a crazy interpretation of the poll results since it is at the heart of the public debate: whether the proposal ties the hands of this and the next president. Indeed, dig deeper in the poll results and you find that 38% of Republicans said they currently support “sending a large number of U.S. ground forces” to combat ISIL. So, in the end perhaps 38% support a broader AUMF, and even more Republicans believe an AUMF should keep the president’s options open. I should note that Sargent may be correct that there is a disjuncture between Republican voters and the GOP lawmakers. But this poll doesn’t give us that information.  [Greg and I had an exchange on Twitter about our interpretations of the survey.]

An important aside: since the President’s proposal may be perceived as limiting his authority compared to the status quo (due to the sunset clause and troop limits), it is even more difficult to interpret these results. For example, the large number of Democrats who support the proposal may be due to their belief that it adds constraints on the ongoing and future military operations against ISIL. (That said, I agree with Marty Lederman’s post and Jack Goldsmith’s post that the proposed AUMF, as currently drafted, would expand, not contract, the President’s authority.)

Finally, let’s consider the depth of Americans’ support for sending US forces to combat ISIL. Here all I want to do is refer back to two earlier posts (here and here) that I wrote on weaknesses in these types of surveys. Drawing on social science research in the academy that analyzes such surveys, I highlighted questions about the depth of public support for military operations against ISIL. A key finding suggests that American public support for such military operations declines greatly in the event of civilian casualties. (Indeed, one of those social science studies involves a critique of polls like NBC News/Marist for failing to include the element of civilian casualties in the survey questions.) The point for this morning’s poll is that Americans might say they support these efforts to use military force today, but if the conflict begins to incur greater civilian casualties that support might plummet.