An expansion of the tropical belt has been extensively reported in observations, reanalyses, and climate model simulations, but there is a great deal of uncertainty in estimates of the rate of widening as different diagnostics give a wide range of results. This study critically examines robust diagnostics for the width of the tropical belt to explore their seasonality, interannual variability, and multi-decadal trends. These diagnostics are motivated by an exploration of two simple models of the Hadley circulation and subtropical jets. The width based on the latitudes of the maximum tropospheric dry bulk static stability, measuring the difference in potential temperature between the tropopause and the surface, is found to be closely coupled to the width based on the subtropical jet cores on all timescales. In contrast, the tropical belt width and Northern Hemisphere edge latitudes based on the latitudes at which the vertically-averaged streamfunction vanishes, a measure of the Hadley circulation's poleward edges, lags those of the other diagnostics by approximately one month. The tropical belt width varies by up to ten degrees latitude among the diagnostics, with trends in the tropical belt width ranging from -0.5 to 2.0 degrees per decade over the 1979-2012 period. Nevertheless, in agreement with previous studies nearly all diagnostics exhibit a widening trend, although the streamfunction diagnostic exhibits a significantly stronger widening than either the jet or dry bulk stability diagnostics. Finally, GPS radio occultation observations are used to assess the ability of the reanalyses to reproduce the tropical belt width, finding that they better situate the latitudes of maximum bulk stability versus those of the subtropical jets.