>Isn't there a problem with that way of accounting since the compensation
>and production come in different periods, perhaps years apart?
Ok, just spoke with a guy at the Bureau of Economic Analysis, who said: 1) proceeds of exercised options are included in wage & salary income - they don't know how much, since it just blends in with the rest; 2) the timing problem probably isn't important, since the exercises are more or less continuous rather than "lumpy; and 3) the option problem probably has little to do with the large statistical discrepancy of the last few years. The stat discrepancy is the difference between income & product, which conceptually are supposed to be equal, since income is earned in production, but aren't in practice, since the numbers are estimated separately; product data is considered more reliable than income data, but income has been exceeding product by a large amount for the last several years, and no one at BEA knows exactly why.
Doug