Read Greg Ip's WSJ piece talking about his new book.  He appears to believe in a "lulling hypothesis" that ignorant urbanites will continue to live in increasingly risky cities below rising sea level (think Miami and NYC) because they have been lulled by big government mega projects into feeling safe.  This argument suggests that public investment in protection crowds out private self protection. Kousky et. al. discuss this point in their 2005 paper and Leah Boustan and Paul Rhode and I discuss this point in this paper.   I like his emphasis on spatial moral hazard but I believe that he is over-stating his point.

Ip is working on ideas that I sketched out five years ago (I would like a citation from him!). For the shortest (and ungated version of my thinking) read this 5 year old document I wrote for Vox Eu.    Superstar cities compete against each other.  As has been shown in the IO literature, rival products have strong incentives to point out the flaws of their rivals. If Manhattan is at risk while Chicago is at less risk, Chicago's land owners and Mayor has strong incentives to send out this news.  Ip, implicitly has a model of PT Barnum's "suckers" in mind. In this age of Twitter and the Internet, I reject this logic. We know that we don't know what risks climate change poses to coastal areas and we will build in more resilience and flexibility (i.e options) so that we don't suffer as much from the next Hurricane Sandy.

The future mayors of NYC have a strong incentive to invest in the city's resilience. There will be a brain drain if NYC does suffer significantly from sea level rise. The only ugly scenario that I foresee would occur if there is a large gap between perception and reality of risk. Suppose that future Don Trumps convince the naive public that Southern Manhattan is safe when in reality it isn't. In this case, people will move to the area and "let their guard down" and not be ready when disaster strikes.

There are two pathways to optimism here;

1. rational expectations --- that perceptions of risk are strongly correlated with actual risk

2. Robust decision rules and risk aversion --- people don't like risk and know that they don't know the emerging risks and thus build in options to protect themselves.

Does IP reject #1 and #2 for what % of the population?