Editor’s Note: CNN analyst Juliette Kayyem is the author of the best-seller “Security Mom: My Life Protecting the Home and Homeland.” She is a professor at Harvard’s Kennedy School, a former assistant secretary in the Department of Homeland Security in the Obama administration and CEO of Zemcar. The opinions expressed in this commentary are hers.

(CNN) —  

4,645 dead. That is the new estimate of the death toll in Puerto Rico after last year’s Hurricane Maria, according to a Harvard University study out this week.

It is no slight uptick from the official number that the United States has been using, based on the Puerto Rico government’s estimate. That number is 64 – which was suspiciously low for a still-recovering island that suffered under a delayed and inadequate response from the federal government.

It is important to note that the Harvard study’s figure is more than twice the death toll from Hurricane Katrina – which happened during the George W. Bush presidency – and puts to rest Donald Trump’s boastful claim that during Maria few had died. For a President who likes to put much in competitive terms, it’s not a comparison he can win now. He is indeed “the best” at something: The single most deadly natural disaster in modern America happened on his watch.

There is something beyond maddening and tragic here; lowballing the mortality means we will not have an accounting that can help us to do better the next time. If we don’t accurately assess what went wrong in the response to Maria, we set ourselves up for another disaster.

Maria was the third major hurricane that season – one in which the response was shamefully delayed and disorganized. Or as NPR noted, “The federal government had three times as many people on the ground in Texas (after Harvey), and twice as many in Florida (after Irma), as it did in Puerto Rico.” The already weak infrastructure on the island crumbled.

What the Harvard study exposes is more than just a number discrepancy. This is, in the end, not about numbers. It is a moral failure of such epic proportions that even someone like me – somewhat hardened by a career in the homeland security field – can’t quite contain the emotions. How could the Trump administration get it so wrong? The answer is that it did not put a priority on getting it right.

I have a great confidence in the career experts working in emergency management. But the political decision to simply ignore, to brush away, what the number of fatalities could actually be is one of negligence and hubris. It is as if the thinking went: If no one notices how many died, then the administration can’t be blamed for their failed response.

Their priorities were not about unearthing the truth; it is absurd that it would take an outside research team to tell the US government what it should already know.

And at bottom, the United States can’t respond in the future if it can’t answer this and face what happened there.

The Harvard researchers arrived at their number by using an on-the-ground, community-based approach – knocking on doors and talking to citizens about who they may have lost from the storm and from the desperate aftermath, where there was a lack food, medicine and water.

Indeed, some analysts of the conditions in Puerto Rico worry the loss of life might be even greater. In a news conference Tuesday, the head of Puerto Rico’s public safety department said the government did not have reason to question the researchers’ new estimates. And according to The New York Times, “The researchers say their estimate, published Tuesday in The New England Journal of Medicine, remains imprecise, with more definitive studies still to come.”

And the outrage and sadness of this is made worse by the fact that a promised US-commissioned study of the death toll has been delayed. As a result, treating those victims with the dignity they deserve is also delayed. But there is also a deadly serious, practical reason for getting this right.

In disaster management, there are five stages to a crisis: prevention, protection, response, recovery and resiliency. The “boom” (crisis event) occurs between the protection and response stages, and is why people in the field talk about the left (before) and right (after) of “boom.”

Each of these stages has its own focus and policies. But it is really the last, resiliency, that sets the foundation for them all. Resilient systems are better equipped to handle the next boom. Ensuring that we build resiliency after a disaster actually makes us less prone to a disaster the next time.

Resiliency is more than a mindset or a mood; it describes a series of investments undertaken to make a system better than before. Because there will be more hurricanes (and terror attacks and flooding and pandemics, etc.).

A major component of a resilient system is one that takes “lessons learned” seriously. Evaluating this is similar to preparing a military “after action” report: It should provide an honest assessment for leaders of what worked and didn’t work. It assumes the entity that has suffered a harm or disruption has every incentive to do better the next time.

In other words, a resilient system “assumes the boom,” and learns from it so it can adapt for next time. We need to assume the boom. There will be more hurricanes. A system that can learn from what went wrong is both more resilient and able to teach other jurisdictions how to be so.

A little realism here: That people die in a disaster is a given. It’s the nature of disasters. But how they die is a different analysis all together. It is government’s obligation to learn this “how” to protect life better the next time. For example, nearly 100 people died in the 1978 blizzard in New England.

Most did not die from snow, but from injuries or mishaps related to the accumulation of snow and difficulties of transport. Many died because the storm came in so fast that people were not told to get off roads, cars stalled and eventually passengers were overcome by carbon monoxide poisoning. How they died matters and is why today a common response to a blizzard is for states to declare driving bans so motorists stay off roads.

The Harvard study found that it was mostly disruption to health care services that contributed to the deaths. The elderly and those with illnesses lost basic utility services, access to medical facilities and medications resulting in their deaths. In other words, they didn’t die from the rain or storm.

Who lives and dies is not a matter of luck. And if we could embrace a rigorous “lessons learned” effort, then more lives could be saved for the next time (or wherever the boom happens).

It’s hard to build resiliency when the official death toll is so low. It makes it seem that there was no great tragedy, but it also keeps us from learning how these Americans died: dehydration, illness, food or what? And it matters because what we learn will change the way we deliver services or think about the next disaster.

At the margins, there may be some debates about “causation” for some of these deaths in Puerto Rico – but not when the numbers are deep into the thousands. The United States has apparently decided not only that those lost souls do not count as victims of the hurricane, but that it very well may make the same response mistakes again.

And again. And again. Hurricane season starts Friday.