The periods cannot be very long in, for example, years, because the LRAS theory demands an initial total output equilibrium followed by a return to that output after a disequilibrating experience.
However, the LRAS argument involves other changes either explicitly on its curve or directly affecting it but not explicitly represented.
That empirical resolution will also critique the LRAS concepts regarding real wages over their "long run.
Examination of the post-WWII data record in Table 1, organized into the nine selected four-year ("long-run") Periods with two biennia each, reveals that the LRAS curve argument is empirically supported in most, though not all, cases with respect to the initial biennia.
This empirical record factually resolves the LRAS theory's real wages ambiguity for the first biennium.
The crucial output data fail to support the LRAS curve, even when the "long run" is absurdly defined as three years of clock time
2) aggregate hours typically fail to fall as the LRAS theory claims.
That secular pattern dominated shorter period compensation behavior, making it impossible to integrate the model of the LRAS curve into any meaningful, critical, empirical analysis for the last three Periods in Table 1;
Most importantly, in the strategic subsequent biennia, nominal GNP was up in every case, and unfortunately for the LRAS model, total output was also up in every case, though but slightly in 1956-58 and 1990-92.
Again, the LRAS model, which predicts decreases, falls the empirical test.
Real compensation at the end of the "long run" is supposed, according to the LRAS model, to be the outcome of a conflict between business firms' price increases and labors' nominal wage raising.
It is fortunate that growth theory has rejected any reliance upon the LRAS curve.