The area near the junction of Arkansas, Missouri, Tennessee, and Kentucky had a set of bad earthquakes in 1811 and 1812.  Further research has shown that similar earthquakes had recurred several times in the previous 2000 years.  However the risk has been difficult to evaluate, as the ground is not measurably deforming and the fault system in the area is poorly understood.

High earthquake danger has been estimated by the USGS for good reasons - the history of frequent recent earthquakes, the ability of ground to the East to transmit strong motions efficiently, and the vulnerability of eastern cities.  So the hazard near New Madrid stands out as a red bull's eye on the national map.

No danger is forseen by Northwestern Prof. Seth Stein in his new book Disaster Deferred and numerous radio interviews and talks for the highly speculative, although conceivably correct, reason that in the long run, faults must reload by tectonic motions, and New Madrid has deformed precious little in the last decade.  He and his former student Mike Wysession go so far as suggest the USGS has lost objectivity because high hazard results in more attention and work for them.

As a result, some emergency managers and legislators have felt free to choose the most convenient assessments to favor.  This issue involved lots of money as well as public safety because earthquake safe buildings cost a few percent more, which adds up to many tens of millions of dollars or more.  Even more seriously, large projects may be located on the basis of safety, and even mild earthquake danger may steer the construction toward less shaky ground.