Based on dividend capitalization, Holt presents a model for determining the appropriate
relative PE ratio for a growth stock as compared to other stocks, whereas Malkiel presents an independent approach for valuing a growth stock relative to a non-growth stock.
We also included company size, the time and absolute PE interaction, and the time and relative PE interaction as control variables.
[Z.sub.2ij] = the relative PE of accountant i in time block j (See Lazear 1999) (-2 [less than or equal to] [Z.sub.2ij] [greater than or equal to] 2): the relative PE is measured as the difference between accountant I's performance evaluation in time block j and the average PE of all accountants at a given rank in a given company in time block j (i.e., [Z.sub.ij] - Z), (4)
[Z.sub.5ij] = the interaction of time to promotion and relative PE, computed like [Z.sub.4ij], but substituting relative for absolute performance; and
We used the relative PE measures at the staff and senior rank (i.e., [Z.sub.2ij] and [Z.sub.ij]) to test H5.
In our archival data, a larger increase in promotion probabilities for absolute PEs, and a smaller increase in promotion probabilities with relative PEs, will provide evidence that is consistent with a work-life based promotion process.
