The human costs of the flooding in Texas are heart-wrenching, as they are in Bangladesh and India, where many more have perished.
The disaster poses a stark reminder of the importance of a functional, amply funded government sector. One hopes (against hope... and the current reality) that Harvey will move Washington's discourse from shutdown, debt defaults and wasteful, regressive tax cuts to good governance, the absence of which is at the heart of the Houston disaster.
The disaster is in no small part a function of four intersecting problems: climate change, bad price signals, hyperbolic discounting and stagnant priors. True, those last three sound annoyingly technical, and to some extent they're just obscure ways of saying humans are programmed not to give a damn about the future. But let me briefly explain each in turn; taken together, they clarify the policy solution.
The role of climate change in exacerbating Harvey's impact has been thoroughly discussed, and the consensus was well summarised by two scientists writing the following in the Times: "Climate change doesn't cause extreme events. It amplifies them. And in any weather-related calamity, our susceptibility to harm is, at its root, constructed by ourselves." What they mean by that last bit is where these other problems come in.
PRICE SIGNALS Price signals are among the most important organising principles of capitalist economies. Scarcity, supply and demand, the extent of risk - we count on prices to convey critical information about these factors, and when they fail to do so, distortions and imbalances occur that, in the case of Harvey, were lethal.
Insurance against flood risks was systemically underpriced for years, encouraging developers to build homes in areas susceptible to hurricane-induced floods (underpriced risk was also at the heart of the housing bubble and Great Recession).
Hyperbolic discounting is a concept from behavioural economics which maintains that our innate impatience is such that we'd rather get less today than more tomorrow. We're generally weak when it comes to delaying gratification, even when to do so would make us better off down the road. And sometimes, we so thoroughly "discount" the future that a nice house in a vulnerable flood plain becomes simply a nice house, full stop.
Stagnant priors imply an unwillingness to update important probabilities based on new information.
Probability theory teaches that additional information can refine our probabilities, causing us to "update our priors". If I drew a 10 from a deck of cards, you'd correctly note that the likelihood of that draw was 4/52, as there are four 10s in the deck. If I then told you that my 10 was red, you'd update your prior probability to 2/52.
In this context, flexible priors mean that if you start having 500-year floods every few years, you update your probability about the likelihood of those floods' occurrence.
The extent to which these tendencies collide is thoroughly covered in a Politico article by Mr Michael Grunwald, "How Washington Made Harvey Worse". The Aug 29 piece cites a report on how the government's flood insurance programme was perversely "encouraging Americans to build and rebuild in flood-prone areas... The most egregious example was a home that had flooded 16 times in 18 years, netting its owners more than US$800,000 even though it was valued at less than $115,000".
It's a clear example of distortionary price signals: the premiums for this federally subsidised insurance systematically underpriced flood risk.
And where was that home? You got it: Houston.
But here's the real clincher: That report was written 19 years ago.
"Houston's problem," Mr Grunwald writes, "was runaway development in flood-prone areas, accelerated by heavily subsidised federal flood insurance."
The development piece is equally implicated as the city recklessly paved over its natural water-absorptive capacity ("anything goes" zoning is also implicated). The urgency of this near-term profit motive just raised the factor by which city government and residents discounted their future.
The question is how could the politics fail so repetitively, blatantly, and fatefully?
In fact, in 2012, a bipartisan coalition in Congress passed a Bill intended to fix the broken price signals, so that premiums to the government's flood insurance programme would more accurately reflect the risks. Then, in 2014, "after an uproar from coastal and riverfront communities, Congress reversed itself in equally bipartisan fashion". (Note: Because of the potential for pervasive, catastrophic damage, private insurers generally won't touch flood risk, which also tells you something about the problem.)
I asked Mr Grunwald why America's politics are incapable of addressing these known problems. Is it really as simple as there are no grown-ups willing to say: "Sorry, but we can't keep getting this wrong?"
"Part of the problem," he told me, "is federal subsidies that incentivise building and rebuilding in flood- prone areas. Another part is the toxic special-interest politics that has made it impossible to reverse those incentives.
"Bureaucratic inertia makes everything worse, and helps explain why the flood maps that are supposed to identify dangerous areas are so absurdly obsolete. All of these problems are intensified because flood-plain development is a land-use problem, and Americans go nuts over federal intervention in land-use problems.
"Then there's the unbelievably powerful real estate industry - Realtors, home builders, mortgage bankers. They're not like Wall Street or Big Oil - they're big in every state and every congressional district."
This leaves us with a huge but transparent problem, one that can be solved only by functional, forward-looking, fact-based governance. In fact, one of the main reasons we form governments is to enact and enforce regulations that prevent us from doing what we're wired to do: ignore and underprice future risks. The increasingly existential threat of climate change only underscores the importance of getting these regulations right.
It is in this regard that "our susceptibility to harm is... constructed by ourselves".
The necessary reconstruction will require new zoning, up-to-date flood mapping, accurate price signals (reduced subsidies), and aggressive, global efforts to both adapt to and mitigate climate change. I realise that's miles away from today's reality, but that just tells you what you already knew: our current reality is unsustainable.
• The writer, a former chief economist to US Vice-President Joe Biden, is a senior fellow at the Centre on Budget and Policy Priorities and author of the new book The Reconnection Agenda: Reuniting Growth And Prosperity.