On Tuesday, I mocked the Hurricane Katrina/Obamacare comparisons. Hurricane Katrina, after all, killed more than 1,800 people. Obamacare's Web site isn't working well yet.

President Obama sells Obamacare. (Win McNamee/Getty Images)

Ron Fournier, among others, defends the comparison, arguing that of course the crises themselves aren't similar. The analogy is really about "the ways that Bush and Obama handled their crises, and [how] those actions changed the public's view of their presidencies."

But that's precisely the point: The political fallout is driven by the nature of the crises. Or, to put it more starkly, reality drives politics — politics doesn't drive reality.

Start with the basic premise of the analogy: Did Hurricane Katrina really doom George W. Bush's presidency? Here's Gallup's chart of President Bush's approval ratings. Can you pick out Katrina?

Hurricane Katrina hit in August of 2005. To be specific, it hit from August 23rd to August 30th. That's a hard period to spy on this chart.

Gallup's fuller results make this crystal clear: In early 2005, Bush's approval ratings were above 50 percent. His slide began mid-year. By July 28, 2005 — so, about a month before Katrina — Bush's approval rating was down to 44 percent. On Aug. 25 — so, during Katrina —  it was 40 percent. On Sept. 8, it was back to 46 percent. On Sept. 28, it was still 46 percent.

Bush continued to drop through the fall, and come 2006, he fell into the 30s and stayed there. But the polls don't support the case that Katrina was the pivotal event.

Instead, Bush's descent was driven by Iraq. The best place to see this is in polls of Americans' top priorities. Early in September 2005 — at the same moment Bush's numbers are rebounding — "Hurricane relief" tops the polls. But in 2006, Iraq takes over and basically never lets go. Which makes sense: Hurricane Katrina was a one-time tragedy, and for most Americans outside New Orleans, by 2006, it was over. Iraq was an ongoing disaster.

The reason comparisons between Obamacare and Hurricane Katrina don't make sense is that they are fundamentally different kinds of crises. Katrina was a fast-moving catastrophe. Obamacare is an ongoing program. The failure to effectively respond to Katrina was irreversible: The men, women and children relying on that response were dead. The failure to launch Obamacare with a working Web site is eminently reversible: The Web site can be fixed.

Fournier blurs this a bit by adding Iraq to his comparison, even though, as his examples make clear, he's actually responding to a conversation about Katrina. But politically, Iraq is a stronger comparison: It's an ongoing, high-profile event that can get either better or worse over time.

But here, too, reality drives politics, and not the other way around. The problem with Iraq was that Iraq was genuinely a disaster. Fournier blames "mismanagement," "tone-deaf responses" and "defensive, insular advisers" for the political aftermath, but that's putting the politics first. There's no amount of messaging, White House staffing  or even management* that could've prevented Iraq's bloody civil war. Iraq wasn't unpopular because Bush failed to lead. It was unpopular because it was a catastrophe that left thousands of Americans and potentially more than 100,000 Iraqis dead.

One way or the other, HealthCare.Gov isn't going to be that kind of a disaster. But it could certainly prove to be its own kind of disaster if, unexpectedly, the digital architecture proves unsalvageable and no reasonable workaround is created. But again, the politics here will be driven by the reality. If the policy continues to fail, then there's nothing the White House can do to keep from being dragged down. Conversely, if the Web site is fixed come mid-December, and the policy begins working pretty well, then there's no amount of Republican messaging that can make it a failure.

While I was writing this post, a reader e-mailed:

I keep hearing about how bad the website is doing and I must say I am getting tired of hearing this and here's why:

I did have problems for the first two weeks, but since then I was able to view all the plans that were available in the State of Texas and was able to narrow it down to 3 plans and once the subsidies kick in I will pay on the low side $50.44 to $63.88 for an all states plan. This is half what I would have paid if I would have stayed with Home Depots junk policy that they offered to part time employee's. That would have cost me $118.23 a month. for far less coverage.

Ultimately, the success of HealthCare.Gov relies on this story becoming the norm for people trying to sign up, rather than the exception. Political writers want to call Obamacare President Obama's Katrina or, pace Fournier, Obama's Iraq, but ultimately this is Obama's Obamacare. And like Katrina and Iraq and Medicare Part D and pretty much everything else that actually touches the lives of Americans, its political success will rely on its actual success.

*This is more controversial, of course, but I subscribe to the argument Sam Rosenfeld and Matt Yglesias laid out in "The Incompetence Dodge": Iraq was a disaster because the underlying idea was bad. Better management of the war would've helped, but it wouldn't have prevented Iraq's bloody civil war or the deadly insurgency.