In my new book, I show that more than 90% of federal disaster funds went to roads, utilities and other standard infrastructure in the early 21st century, and less than 10% to libraries, parks, schools, places of worship and community centers, where neighbors check on one another and organize recovery.

Federal aid may go to rebuild some of these spaces, but it rarely pays to keep them open and running when communities need them most.

In another study, I found that Japanese communities with higher density of social infrastructure had half the number of deaths from a catastrophic tsunami in March 2011 compared to similar communities with fewer gathering spaces.

In my assessment, as federal aid becomes less predictable, these kinds of places people frequent often will have to carry us into the future. The communities that recover best from the next disaster may be the ones that kept their libraries open, their civic halls full of volunteers and their parks full of residents.