Although the SEC staff has publicly stated that the clawback provision is the most complex of the remaining Dodd-Frank governance rulemaking, the most controversial provision appears to be the requirement to disclose the ratio between the CEO total compensation and the employee median.  Given the specific and prescriptive nature of the legislation, interested parties have been struggling for some time to come up with a workable solution, while some continue to argue for outright repeal.

In August, the AFL-CIO suggested that the SEC could permit issuers to comply through the use of statistical sampling, with SEC guidance on how to construct the sample methodology, such as sample size and specific confidence levels, as well as allowing issuers to identify the median employee based solely on cash compensation.  Certain SEC staff members have publicly declared the idea "interesting" and worth considering as they develop proposed rules.  The SEC website still indicates that proposed rules are planned before the end of 2011.

Now the Center for Executive Compensation, which represents the senior HR representatives of more than 325 of the largest US corporations, has responded with its own letter to the SEC.  After conducting a survey of its members, the Center concluded that statistical sampling would be extremely difficult for most global companies to implement, primarily due to the wide variability in pay practices and recordkeeping.  More than three-quarters of respondents have over 10,000 employees globally, with significant populations located in more than 10 countries.  Almost half said that manual calculations would be the predominant approach and that it would take at least three months to calculate the ratio, so that proxy deadlines may not be met.

The primary concerns stem from the complexity of data aggregation given the fact that most companies, especially large multinationals, do not maintain a centralized list of employees that is linked to their compensation information.  Rather, since information about each employee is maintained in the separate country and business unit, statistical sampling would not avoid the need to develop a list that first identified all employees and then identified and calculated all elements of compensation needed for sampling purposes.  Some of the specific issuer responses noted that sampling would require examining local payrolls in over 30 countries or as many as 100 different payroll systems, and countless vendors.  Other concerns raised included the implications of currency values, tax impacts and privacy laws.  Issuers also questioned whether total cash compensation is an appropriate proxy for total compensation for non-US employees, as "in kind" contributions make up a substantial part of compensation in certain parts of the world for tax and cultural reasons.

"Issuers appreciate the SEC staff's efforts to find a flexible approach for a very challenging requirement, but unfortunately statistical sampling would neither eliminate, nor reduce the significant difficulties associated with gathering the data necessary to accurately identify the median employee based on compensation.  For large multinational companies with multiple operating units like J&J, we can already see that this data gathering exercise will present major administrative burdens.  What we need is a practical solution that does not place additional strains on companies' resources in these tough economic times," commented Doug Chia, Corporate Secretary at Johnson & Johnson, which has 117,000 employees working in over 250 operating companies worldwide.

The Center concludes its letter by recommending that the SEC consider the use of estimates similar to median employee pay that are already publicly available, such as to employee earnings maintained by the Department of Labor's Bureau of Labor Statistics, which can be customized for a company's primary industries.