The 32-month American-led air campaign against the so-called Islamic State, which was initiated by the Obama Administration and has continued under Donald Trump, has only recently captured the United States public’s attention, with reports of a large increase in collateral damage and civilian casualties. According to careful analysis by the monitoring group Airwars, there were hundreds of civilian fatalities tied to coalition activity last month alone, a number comparable to fatalities reported during the height of the Russian bombing campaign in Syria last year. In total, the research organisation estimates that coalition airstrikes have killed at least 2,978 civilians.

It would be prudent for the public to approach these reports, and any speculation about the underlying causes, with a degree of caution and healthy scepticism. But that may be impossible unless policymakers in Washington, especially Congress, take their own oversight role more seriously.

The truth is that as an outside observer, it is impossible to know why there has been a reported increase in civilian casualties recently or what accounting or methodological standards would explain the huge divergence between estimates by Airwars and US Central Command, which assesses that “at least 229 civilians have been unintentionally killed” by all 19,300 coalition strikes conducted since the war began in August 2014. Outsiders are not allowed inside the Combined Air Operations Centre, the government nerve centre that monitors and directs the air assets supporting the war. Without witnessing and evaluating the individual behaviours and organisational culture of those involved, you cannot truly appreciate whether the US military’s policies to protect civilians are translated into actual practice.

We also cannot know the sources and analyses of intelligence for particular strikes, nor are we privy to targeting guidance, command arrangements, collateral damage estimation methods, or weaponeering, which, cumulatively, would be required to understand the root causes of any civilian casualty incident and to assign blame for those who failed to follow the rules under which they were operating. The command arrangements would be particularly meaningful for any evaluation, as they were for the March 17 strike in Mosul, which killed more than 100 people. That strike was actually a “dual-key” offensive in which an Iraqi commander had to simultaneously authorise the attack along with his American counterpart.

Similarly, be wary of claims by civilian and military officials, like Secretary of Defence James Mattis, who said: “We go out of our way to always do everything humanly possible to reduce the loss of life or injury among innocent people.”

Of course, neither he nor any other official is omniscient and could possibly be sure about such a categorical assertion. The US military is extremely transparent about how it should theoretically conduct military operations while being needlessly secretive about whether the many tactics, techniques and procedures are actually and faithfully followed. Military commanders proclaim constantly: “You wouldn’t believe all the steps we take before we approve a strike.”

As if the sheer volume of processes was any indication of their necessity or effectiveness for preventing civilian harm.

As I have documented by assessing military investigations into civilian casualty incidents, people in the chain of command are often deployed without adequate training, skip critical steps — such as updating commanders when the environment changes — do not allow for sufficient time to review the intelligence, experience pressures from commanders to request more and more airstrikes, and make any number of other avoidable human errors. Moreover, if the Pentagon actually pursued “everything humanly possible” to protect innocent non-combatants, it would at least consider halting strikes completely. Of course, that is not a plausible option given the initial framing assumptions that are widely accepted within Washington and that continue to inform the war.

In late September 2016, I offered a specific recommendation for how the coalition could potentially reduce civilian harm:

When time is available, the underlying intelligence and operational outcome of significant strikes should be subjected to independent assessments by “red teams” comprising analysts outside of the target development process. Those who plan and conduct military operations cannot review and critique their own efforts.

Twelve days before this was published, American, British, Danish, and Australian aircraft conducted 37 airstrikes against a Syrian military convoy south of Deir Ezzor airfield. The coalition had been monitoring the target for two to three days, during which time the analysts, lawyers, commanders and others within the CAOC “positively identified” the targeted ground forces — heavily armed and wearing a mix of civilian and military clothing, with no flags or insignias — as being Islamic State fighters. This conclusion turned out to be false. After 27 minutes of bombing, Russian military officials informed their US and coalition counterparts over the deconfliction hotline that the strikes were actually hitting Syrian regime forces.

An investigation into the airstrikes by Brigadier General Richard Coe “found no evidence of misconduct”, yet identified multiple process, information and communications errors that resulted in the unintentional airstrikes. These include “several breakdowns in the flow of information into and around the CAOC [that] led to a number of misconceptions”, as well as human factors: confirmation bias, task saturation and target fixation.

After a brief, redacted version of the investigation was published, Coe conducted a press briefing. He revealed that during the target development process, an intelligence analyst watching the forces over full-motion video typed a message into a shared chat room: “What we’re looking at can’t possibly be Isil,” using an alternate acronym for the Islamic State. According to Coe, that warning “was not pushed to a larger group, nor to the final decision-makers because the people were already certain that the target was fully validated and there was no other information to be considered that overrode that input”.

When the investigation results were released, a reporter asked Coe: “You’re saying there was no red team? Or that the red team process wasn’t used in this case?” To which he replied: “There was no one ... designated to play that role. Each person is expected to do that on their own and then in the process funnel up the pros and cons to decision-makers.”

This is a truly dangerous assumption for any organisation to make: believing that analysts immersed in finding targets can detach themselves from the process to weigh the pros and cons, and then be expected to “voice up” their dissenting viewpoint that runs counter to everyone else’s. To my knowledge, there is still no red team within the CAOC designated to review the underlying intelligence and general soundness of coalition airstrikes in Iraq and Syria.

As someone who relies on open-source information because I lack access to classified information, I cannot fully appreciate the safeguards involved in targeting processes; nor fully trust insiders’ claims. This is why congressional oversight committees, empowered with both access and information, should act on the public’s behalf to investigate and evaluate the civilian casualty prevention policies and processes being used at present in the bombing campaign. Unfortunately, in public hearings policymakers have abrogated this responsibility and expressed little interest in fulfilling their role. In fact, most of the questions asked of uniformed officers inquire whether they want looser rules of engagement in attacking the Islamic State.

Members of Congress are free to adopt a hands-off approach to yet another war in the Middle East. But in doing so, they confirm that the appearance of adhering to the standards of necessity and discrimination in the use of force is as important as the actual attacks themselves. The result will likely be the unnecessary deaths of hundreds more Iraqi and Syrian civilians, which the American public, and their constituents, should not let them forget.

• Micah Zenko is a senior fellow with the Centre for Preventive Action at the Council on Foreign Relations and is the author of Red Team: How to Succeed by Thinking Like the Enemy