We study management turnover for the top five executives in a sample of 443 large firms from 1993 through 1998. The rate of forced turnover for non-CEOs is at least as great as that for CEOs, but the sensitivity of turnover to firm performance is smaller for non-CEOs. The probability that a non-CEO leaves office is elevated around CEO dismissals, particularly when the replacement CEO is an outsider. Many dismissed executives obtain new employment, but on average their new positions are significantly inferior to their prior jobs. Labor market outcomes are related to past compensation and the circumstances around departure.

We find that a new compensation disclosure item on expected payouts from performance-based stock grants reveals unique information regarding future firm performance. Extracting inferred performance expectations from the disclosures, we find that firms disclosing the highest expected grant payout significantly outperform in ROA, Q, sales growth, and profit margin over the next two years, while those disclosing the lowest expected payout underperform. The embedded information is not captured by other information channels, such as managerial earnings guidance, 10-K sentiment, insider selling activities, unexplained CEO pay, and analyst forecasts. Investors and analysts do not fully incorporate the information and are later surprised around earnings announcement days. A portfolio that buys firms with the highest performance expectation and shorts firms with the lowest expectation earns significantly positive abnormal returns. Our findings suggest that the enhanced compensation disclosure contains valuable information, but investors underreact to information that is difficult to collect and process.