Why Government Salary Data Is Conservative

Salary data from the U.S. government’s Occupational Employment Statistics survey is useful and free, but it is considered to be conservatively low for some business purposes due to a number of reasons.

Wage ceiling: The United States Bureau of Labor Statistics OES survey is more census than traditional survey.  It collects wage data as a head count by pay range within each of 800 broad occupational categories.  For example, the number of incumbent Statisticians paid between $47,320 and $59,799 (the current band for Range F) will be requested, but the top pay band (Range L) is open-ended, defined as “$90 per hour and above” or $187,200 per year. Yet, in actuality, some employees receive much higher pay.  Therefore, high-end outliers do not exist in the OES database despite their occurrence in the real world.  Restricting the highest rate recorded to the low-range bottom of an interval segment would reduce averages for executive pay below the salary rates actually paid, so other methods are used to impute averages or means.

Time frame:  This survey captures data on one-third of a sampling of established American employers each year. In total, it covers 1.2 million establishments listed in State Unemployment Insurance (UI) files.  Updating 400,000 entries once every 3 years means that upsurges in competitive rates can be missed. This makes the raw data to which OES survey applies standardized update estimates considerably antiquated compared to almost any annual private survey.

Organizations surveyed: Only surveying employers on UI  lists eliminates new high-risk start-ups that offer premium pay but have not yet turned up to be surveyed once every three years. Instead, the survey covers larger long-established firms that tend to pay more modest guaranteed base pay in light of the greater total reward value of their benefits, perks, stability of employment, and career progression options.

Limited data variables: This survey is conducted across 335 broad industries. It does not track significant factors that cause variances in pay such as employer size and (in many cases) location, type of business and industry subcategory. The OES survey does not offer refined differentiation between for-profit and nonprofit organizations (which tend to pay lower salaries). In addition, it specifically excludes certain industries and occupations (agriculture, fishing, forestry, household workers and pilots, for example) while including government workers (except security forces).

It is a well-known fact that government pay is low. Government service at a professional level almost always involves a sacrifice in terms of current cash earnings.  Managers who leave public employment generally do so to receive private-sector pay that can be orders of magnitude greater than their public paychecks*. All those factors tend to produce conservative survey rates, particularly at upper income levels.

Are you aware of any other reasons?  BLS data can be very useful but must be applied with full knowledge of its limitations.
 
E. James (Jim) Brennan is Senior Associate of ERI Economic Research Institute, the premier publisher of interactive pay and living-cost surveys. Semi-retired after over 40 years in HR corporate and consulting roles throughout the U.S. and Canada, he’s pretty much been there done that (articles, books, speeches, seminars, radio/TV, advisory posts, in-trial expert witness stuff, etc.) and will express his opinion on almost anything.

*For more information, see the article: “Public Sector Pay and Performance.”

Comments

[Orlando Electrical]

Wow..Nice Blog, Nice tips, Great article!, I have recently started a blog, the information you provide on this site has helped me tremendously. Thank you for all of your time & work. :)

[E. James (Jim) Brennan]

Appreciate the feedback.

What good is knowledge if you don't share it? We all benefit by pooling information for mutual benefit.

[Compensation Consultant]

Many employers elect not to report their pay information since they would not use this source as a salary survey themselves. It can be an expensive waste of time for large employers with multiple locations and is often an early budget cut, in my experience.