Quarterly financial reports are an important way for companies to communicate with the capital market, and they clearly affect firms’ access to debt and equity capital. However, the most scarce and sought after resource for many modern, cash-rich companies is not financing but people: human capital. For example, 27 percent of U.S. employees voluntarily changed jobs in 2019, costing firms an estimated $475 billion. For perspective, the total interest expense reported by public companies was $467 billion, indicating that the cost of voluntary employee turnover is comparable to firms’ external cost of debt. Total turnover-related costs are even larger when taking into account the productivity losses from employees who consider leaving but ultimately stay.
Our recent study investigates whether and how firms’ quarterly financial reports affect voluntary turnover by rank-and-file employees. Our central hypothesis and finding are that firms’ earnings announcements (“EAs”) provide employees with information about the value of their current positions, which causes them to update their beliefs and can lead to job searches. For example, EAs provide information about the firm’s financial prospects, which can affect future job security, raises, promotions, implicit commitments, and resume value. It is plausible that rank-and-file employees learn from public reports because they usually work within a discrete function of a firm and do not have access to their employers’ broader accounting systems. Also, even though employees can directly observe some aspects of firm performance, their information is likely incomplete, lacks the credibility of public reports, and lacks the interpretation and context provided by EA conference calls and intermediary coverage.
We identify and investigate short-window changes in job search that can be plausibly attributed to EAs using data from Glassdoor.com. Glassdoor is a website where users post anonymous reviews about their companies in exchange for access to job postings, reviews, and compensation data. This “give-to-get” feature means that many Glassdoor reviews are written by employees who are reconsidering their current employment and beginning to search for other jobs, including both employees who ultimately leave and those who stay. We use weekly counts of new Glassdoor reviews by current employees as a proxy for new job searchers and use abnormal within-quarter changes in review counts to identify changes in search around EAs.
We document significant increases in reviews posted by current employees around EAs, after controlling for firm characteristics (e.g., size, performance, and average job search), macroeconomic or market conditions, and temporal trends in Glassdoor’s user base. The results are consistent with EAs driving greater job search by current employees.
We next examine when and why EAs have a larger effect on voluntary employee turnover. First, we find that EAs have a larger effect when employees have more abundant and attractive outside job opportunities. Second, EAs have a larger effect when within-firm information frictions are greater, i.e., when employees are more geographically dispersed and therefore less able to internally observe accounting signals.
We also examine the role of financial distress and earnings news. While prior studies find that financially distressed firms have trouble attracting and retaining employees, it seems plausible that EAs are a less relevant source of information if distress is easy to observe internally or if employees learn about distress from events such as credit rating downgrades. Across multiple measures of financial distress, we fail to find consistent evidence of EAs having a larger or smaller impact on the employees of distressed firms.
We document similar findings when examining whether the effect of EAs varies with the type of earnings news. EAs with bad news likely cause employees to lower their expectations about the value of their positions and consider moving, at least when the bad news is firm-specific. However, good news EAs can also increase employee job search by increasing the value of the employee’s current position and improving her outside opportunities or by causing her to feel underpaid for her contributions to the company’s success. We find some evidence that EAs have a larger impact on employees when analyst revisions and media tone are negative, but these results do not survive all of our robustness tests.
A final set of analyses investigates whether employees learn from EAs. We find that employees’ business outlook scores are positively correlated with multiple measures of EA news and that the correlation is stronger during EA weeks than earlier in the quarter. These results indicate employees do learn from EAs.
In sum, we provide novel evidence that a critical, non-investor stakeholder group – rank-and-file employees – learn from their employers’ public EAs and that EAs motivate employees to reevaluate their employment and consider leaving. We document that the role of accounting information in labor markets is not limited to the extreme settings of financial distress or restatements examined by prior studies. Instead, we find that accounting information affects employee job search even for firms that are performing well.
Our findings also have implications for human resource management. To counteract the effects of financial reports on job search, managers can consider bolstering their retention efforts around EAs or design financial reporting strategies with employees more in mind.
This post comes to us from professors Ed deHaan at the University of Washington’s Michael G. Foster School of Business, Nan Li at the University of Toronto’s Rotman School of Management, and Frank Zhou at the University of Pennsylvania – The Wharton School. It is based on their recent article, “Financial Reporting and Employee Job Search,” available here.