The United States Constitution does not explicitly state the time zone used.
However, in reality, the Eastern Time Zone which the District of Columbia is in, is followed.
It's likely because it's the local time of the District of Columbia where the inauguration takes place and that D.C. is the ...
@amflare: I would expect that legislation setting the date and time when power would be handed over would have been enacted during George Washington's term of office--a point in history where locations separated by five degrees of longitude would be expected to recognize noon about 20 minutes apart. People writing the legislation would have been aware that noon didn't happen at the same time everywhere, but unless the incoming and outgoing Presidents were at different longitudes, it wouldn't make sense to use anything other than the local time wherever they happened to be, especially since...
[also @NicholasShanks:]...nothing that happened after noon at the President's location would have any effect that could be observed before noon anywhere else. The invention of the telegraph made it possible for someone to find out, before noon, about an event which happened after noon somewhere else, but that still wouldn't imply any reason to use anything other than local time at the incoming/outgoing Presidents' current location. If the Presidents were at different places, the switch-over could be allowed to happen at any moment they agreed upon between the two local noons.
@supercat presumably the actual time zone used is "local civil time in Washington, DC, whatever that might be"; I imagine that in 1945 FDR was sworn in at noon, Eastern War Time, making his third term an hour longer than four years and the 1945-to-1949 term that Truman completed after FDR's death an hour shorter.
@phoog: I'd say local civil time wherever the Presidents happen to be; if they are in different places, then any time that is agreeable to them between the respective noons, or the later noon if no other agreement is reached.
@supercat hm, interesting. If one were unavoidably in a different time zone, though, I somehow doubt that is how the ambiguity would be resolved. I suspect they'd do it at noon DC time. If they were both in some other place, I'd probably give even odds at local noon or noon DC time. I suppose we're unlikely ever to find out.
@supercat it appears that for the earliest inaugurations they were mostly concerned with having it on the correct day (the constitution establishing a term of years, and March 4th taken to be the beginning of Washington's first term). Washington called the session for Adams' inauguration to begin at 10am (in the Senate chamber in Philadelphia), and Jefferson himself "proposed" ... "to take the oath which the Constitution prescribes to the President of the U.S. before enters on the execution of his office, on Wednesday the 4th at twelve oclock in the senate chamber."
@phoog, if there was to be a transfer of power away from the capital, it would be due to crisis or emergency, in which case local time (or time zone) would most decidedly not be a factor. LBJ was sworn in on Executive One while at Love Field in Dallas (at 12:30 CST, buy only by coincidence). Still, a technical inauguration, however.
@200_success: Well... sort of. It would be more accurate to say that standard time had not been invented yet. The result was that (in modern terms) the world observed an unwieldy patchwork of unofficial and nonstandard "time zones." Neighboring cities might be a minute or two off from one another, for example, because there was no systematic timekeeping standard for them to synchronize on. In short, it was a mess.
To elaborate on @amflare's comment, it appears that the USA changed to a Greenwich-based time in 1883. This was confirmed by the International Meridian Conference the following year. In the year 1918 the Standard Time Act came. So in essence, the time zone system was established in its current form long before the 20th Amendment.
@Kevin: Was there any way by which news could travel eastward or westward faster than the Sun? If not, did the concept of "global time" matter to anyone other than astronomers and navigators?
@chrisg LBJ did not take office at the beginning of a constitutionally defined presidential term, so the timing of his oath is not particularly relevant to the meaning of "noon" in the 20th amendment.
@supercat I doubt there was a concept of global time at all, and it certainly didn't matter to navigators, who were interested in the precise difference between their local time and that at some other known reference point, usually their point of origin.
@supercat - If I had to guess, I don't think anyone really understood (or cared about) the concept of global time until the telegraph. Specifically the transatlantic telegraph cable (1858).
@phoog: There may not have been a standard for absolute time, but navigators certainly understood the concept that if they were to sail west 5 degrees in the interval between two noons, everyone in the world would have observed the passage of roughly 24 hours and 20 minutes between the moments the sailors observed those two noons (and that would include the sailors, if they had a good enough chronometer).
@phoog: For most people other than astronomers, traveling east or west would make days slightly shorter or longer, but not by any amount worth worrying about. Navigators, by contrast, would be interested in the exact amount by which days were lengthened or shortened, since being able to measure that would allow them versatility in their routes, making them less vulnerable to ambush (e.g. by pirates). Practical measurement of such time at sea was difficult, but astronomers at land-based observatories could figure out their longitude, given favorable conditions, by observing Jupiter's moons.
@phoog: Since anyone on Earth will see Jupiter's moons get eclipsed at the same time, someone who knows that a particular eclipse will be observed in Greenwich at 11:30pm on Saturday, but observes it at 8:30pm on that day, will know that they are 45 degrees west of Greenwich. That was well understood by 1770. The problem was that conditions on the open ocean were rarely favorable for astronomical observation.
@supercat or they can just observe the position of any celestial object near the ecliptic and compare it to the expected position of that object at their point of origin (which depends on knowing the local time at the point of origin) and then they can calculate how far east or west they've traveled. That method is far less dependant on sensitive conditions such as those required to observe Jupiter's moons. Except in heavy overcast, the object of observation can be the sun, which gives one at least a rough idea of longitude at least once a day.
@phoog: If one knows the time at a location of known longitude, finding one's present longitude is easy. In 1761, there was one clock in existence which could keep time well enough aboard a naval vessel to allow those on vessel to know the time elsewhere. Total eclipses of the Earth's moon are easy to observe, and someone with a clock that had a minute-per-day uncertainty could find out their position within 30 miles or so when such an eclipse occurs, but they only happen a few times per decade. Eclipses of Jupiter's moons are much more frequent.
Discussion on answer by Panda: In whi…
Discussion on answer by Panda: In whi…