. Sample and estimation of abnormal net hiring
Sample
The information for constructing our main variables, such as accounting quality, number of employees, and most firm characteristics, is obtained from COMPUSTAT. Security price and return information is from Center for Research in Security Prices (CRSP). Institutional shareholdings are obtained from Thomson Reuters’ CDA/Spectrum database, and industry unionization data is from the Union Membership and Coverage database.
Our initial sample consists of all 62,165 firm-year observations from 1983 to 2007 with the necessary information in COMPUSTAT and CRSP to estimate the level of expected hiring using (1). We begin the sample in 1983 because that is the first year covered by the Union Membership and Coverage database. The number of observations used to estimate (2) is reduced to 44,861 due to the availability of the various explanatory variables. For example, recall that AQ requires accounting information from years t – 5 to t, as well as at least 20 firms per industry in a given year.
Estimation of abnormal net hiring
To form our measure of abnormal net hiring, we first estimate the expected level based on (1). Descriptive statistics for the variables included in (1) are presented in panel A of Table 1. The