Forgive me for not looking all of this up for you. I responded originally because I strongly suspect there was no peak near the beginning of the Thirties. The crash was in '29, after all. This means the decade would begin at or near a low point, and the average growth rate for the '30's would exaggerate the rate for the relevant cycle.
The concept is not all that squishy. Business cycle peaks are relatively distinct in the data (troughs are less so). The National Bureau of Economic Research issues 'official' designations of when cycles began and ended. Don't ask me who gave them the franchise. Dating the cycles does not appear to have been all that controversial.
The NBER has a web site so the curious might find the dates there. Or Prof. DeLong could enlighten us if he has a spare minute.