(On a side note, the federal government used to have a physical flood model for the Mississippi River. It was literally a miniature, yet massive, outdoor scale model of the river basin’s topography. Water would be sent through the model and they’d see where it flooded. The feds switched over to computer models not because they were more accurate (in fact they weren’t, at least at first), but because the computer model was cheaper and could run simulations much faster. The podcast 99% Invisible has an interesting episode about it.)
Anyway, there are a variety of options to improve the government program, which might even encourage the revival of the private market, at least for supplemental coverage. The NFIP could use up to date maps based on future expected risk, and price based on that risk. Buy out repeatedly flooded properties and return them to nature. Give discounts for flood mitigation, which would eliminate the moral hazard of insuring risky properties. And if people want to pay less they can lobby their local government to implement zoning rules (which Houston lacks) and other projects to reduce flood risk. Localities could be helped in this with federal matching dollars for local mitigation programs, similar to other infrastructure projects (after all, the federal government will pick up part of the tab for floods eventually). Presently, in order to purchase flood insurance the locality the property is in must have a flood mitigation plan, but clearly this has been inadequate and favors better resourced localities. In Maryland, we have something that opponents derisively call the “rain tax”, which taxes property owners based on the impervious square footage of their properties. Such a tax, in theory, internalizes the negative external costs of a property generating storm water runoff, incentivizing private sector mitigation solutions. An insurance mandate may help enlarge the risk pool, but given correlated risks it would also increase payouts.
Undoubtedly these fixes will cause many people to be priced out of owning property in the most risky areas, but that’s the whole point. Less development in such areas would make everyone else safer from floods. And an actuarially sound government run insurance scheme would still charge less than private insurers. Speaking of which, get rid of the role of private insurers in selling, assessing and processing claims.
But of course, asking the feds for money is easier and more immediately rewarding than demanding local flood mitigation programs. And past reforms have been watered down by popular demand. I think the Federal Reserve is a good model for addressing these issues. A lot of government agencies could benefit from such a set up. The fed has been indefinitely authorized by congress and left to make its own decisions about how to achieve legislated goals, often unpopular ones that significantly impact people’s lives. Its board has representation from member branches chosen by member banks as well as political appointees. The fed makes mistakes, but by design nearly all of its mistakes are the result of the general economic consensus being wrong rather than special interests or dumb politicians. Also the fed is audited regularly. So give the NFIP, among others, as strong a mandate and institutional design rather than rely on congressional whims at every five yearly re-authorization.
Alas there is more than economics to everything. Repeatedly flooded properties are clustered in a few states, namely Texas, Louisiana, Florida, New Jersey, and New York. This means policy changes will have large concentrated effects. Just as the costs will not be distributed evenly geographically, they will almost certainly not be distributed evenly across race and income. While property owners tend to be more well off than non-owners, renters will face higher costs if their landlords do. Some properties in greater flood risk, such as beach-front, are pricier for it, while others are made less valuable. An example is New Orleans, it would be perhaps the most heavily impacted community by any changes to the NFIP. The city’s poorest residents live in its most flood prone areas, which are un-coincidentally the most predominantly black. This is generally true for the poor and minorities in cities the world over. This has prompted calls based on environmental and social justice to preserve the affordability of the program in order to preserve poor and minority communities.
In part to address these concerns, Omri Ben-Shahar and Kyle Logue examine the distributional effects of the NFIP and Florida’s state run property insurance program, named Citizens. Crucially, Citizens publishes data on actual rates paid by flood policy holders and the actuarially sound risk-based rates, allowing the researchers to measure the size of subsidies. Unsurprisingly, inland policy holders are overpaying, while the closer you get to the beach the larger the subsidy gets. The authors use data on median home values by zip code and insurance policy caps as proxies for wealth. A property owner cannot take out a policy that pays out more in one flood event than the property is worth, suggesting it is a good proxy for individual level home values. With each proxy the results are generally the same: both the dollar amount and percentage of subsidy increase with wealth. Because Florida’s program is very similar to the NFIP, their results are externally valid.
According to FEMA data from 2006, the average NFIP policy holder pays between 35 – 40% of the full risk premium. Of course those closest to the water pay much less. And 23% of coastal properties with flood insurance are not primary residences, aka mostly vacation homes. Such second, third, and so on homes are worth more on average than flood insured primary residences. The largest subsidies go to rich people who don’t even live in the flood zone. Like all subsidies, while poor people may depend on them, the majority of the benefit goes to the well off. Figure 3, below, from Ben-Shahar and Logue’s paper displays this nicely. The relatively wealthy are the largest subsidy recipients, but as the scatter plot shows, many less affluent people benefit too. While the rich can absorb the increased cost of eliminating the subsidy, or heaven forfend, live in fewer houses, it is the poor who will be priced out of living in certain areas. Again, this detrimental impact on the poor would likely be felt greatest in black neighborhoods of New Orleans.
However, the current program is a horribly inefficient and expensive way to benefit the coastal poor. And such calls ignore both the cost of subsidies born in part by the inland poor and the opportunity cost of wasteful spending. The most obvious change would be to not subsidize non-primary residences. Subsides on insurance could be means tested (which would add administrative costs). Or the cap for payouts could be lower, which would only impact the richest policy holders. And if people want more coverage they could buy private coverage. That is generally how the UK’s healthcare system works: public health insurance for all, but if you want fancier coverage there’s a private market for it. Sure the private flood insurance market collapsed here, but policy changes like this could encourage it to come back. Or my favorite option, eliminate all subsidies everywhere for everyone that can’t be justified on the basis of positive externalities and just give poor people money, preferably in the form of a basic income that phases out as income increases, perhaps modeled off the Refundable Earned Income Tax Credit.
This is similar to arguments I’ve made for free trade. Yes, there will be some concentrated losses for particular communities even though there are net benefits in aggregate. But that means helping those communities cope is easily affordable. There’ll still be displacement and adjustment, but letting the program lumber on as it is will only give more money to the rich at the expense of the poor and increase the number of people, poor and otherwise, who are put in the path of danger by subsidizing unsafe development. Basically, the reasons for putting off the reckoning will only be made worse as time goes on. The cheapest and least disruptive option is to fix the system now.
PS here's a good article about what Louisiana is doing to try to stop sinking into the Gulf.
No comments:
Post a Comment