Immediately after the founding of the United States of America, it was standard practice to refer to the entity in the plural, as a grouping of states rather than a singular body. Then, by about the mid-1800′s, use of “is” or “are” became pretty much interchangeable, though the frequencies at which was more favored varied. You could even use both in the same breath if you wanted to, as displayed in an 1882 case: “[T]he doctrine [of sovereign immunity], if not absolutely limited to cases in which the United States are made defendants by name, is not permitted to interfere with the judicial enforcement of the established rights of plaintiffs when the United States is not a defendant or a necessary party to the suit.”
Today, anyone over the age of four stating that “the United States are located in North America” would be looked at funny, or possibly suspected of taking that whole states rights thing a little too far. Contrary to folklore wisdom, it wasn’t the Civil War that effected the change, but it was rather the result of a gradual semantic shift.
Only in the 1900s did the plural usage fade entirely from Supreme Court opinions. This comports with a contemporaneous account from a commentator in the Yale Law Journal, who observed in 1900 that “the plural use of ‘United States’ is gradually passing, under stress of the ever-increasing sense of unity in the national life.”
“The United States are” has only been seen in a Supreme Court opinion on two occasions since 1901, the last occurrence being in 1935. Unsurprisingly, the sole judicial hold out for the plural form, Justice McReynolds, was one of the Four Horsemen that fought valiantly against the New Deal and the federal tyranny of the commerce clause.