What level of public employee pay and benefits are affordable and appropriate is a difficult but necessary discussion. And missing too often from this discussion is good data on just how much, on average, public employees are currently making. In California, the State controller has made available a database of public employee compensation, organized by agency, that includes every city, county and state worker. The analysis to follow represents the first attempt we know of to extract from the raw data the average pay and benefits for full-time employees of California’s cities, counties, and the state government.
One of the biggest weaknesses inherent in the State controller’s “Government Compensation in California” database is that the summary information provides averages that take into account positions that were part-time, or only occupied by the employee for part of the fiscal year. But the State controller’s compensation website provides downloadable Excel spreadsheets on their “raw export” page that yield sufficient additional information to estimate averages limited to full-time employees. Table 1 shows the difference between averages compiled for all employees, including part-time workers (left three columns) vs. averages compiled for full-time employees, excluding part-time workers (right three columns).
Table 1: Average Compensation, All Employees vs. Full-Time – 2012
As can clearly be seen on the above table, it is extremely misleading to rely on average pay and benefit data that includes part-time and partial-year employees. For example, if a researcher were to click on the State controller’s “Data at a Glance” for Redondo Beach, the “average wages” for a city employee are reported as $47,879 per year, and the “average benefits” (comprised primarily of the employer contribution to pension and health insurance) are reported as $18,203 per year. But if the averages are recalculated to only include full-time employees – a far more representative indication of how much city employees actually make – the average wages increase to $93,809 per year and the average employer-paid benefits increase to $38,197 per year. This results in a total average compensation for full-time employees of $132,006 per year, more than twice as much as the average total compensation of $66,082 as reported on the State Controller’s “Data at a Glance” page for Redondo Beach.
In general, and as shown in the examples provided in Table 1, when limiting the pool of records under analysis to full-time employees, nearly 50% of the records are eliminated and the average pay and benefits increases by nearly 100%.
* * *
METHODS AND ASSUMPTIONS TO ISOLATE FULL-TIME EMPLOYEE RECORDS
The method to remove part-time records from the denominator, in order to develop compensation averages for full-time employees of California’s cities, counties, and state government, using the State controller’s raw data, rests on three assumptions. They are:
(1) A full-time employee would participate in a health insurance plan to which the employer would contribute some portion of the required payment, however minimal.
(2) A full-time employee would participate in a retirement benefit plan, usually a pension, to which the employer would contribute some portion of the required payment, however minimal.
(3) A full time employee would earn an amount in “regular pay” at least equal to the amount specified as the “minimum pay for job classification.”
It is important to note that the pool of full time employees that is isolated using this analysis does not necessarily include all records of full-time employees. The third condition that must be met, for example, that requires a “full time” employee to have earned an amount in “regular pay” at least equal to the amount specified as the “minimum pay for job classification,” will exclude employees who only worked a partial year (typically because during the year they either were hired, retired, or transferred into or out of that job) and therefore earned less than the minimum. But “partial-year” employees must be excluded from the analysis because their lower earnings are not representative of what they would have earned if they’d been in the position the full year.
Another factor worth explaining are end of career payouts of, for example, accrued sick time, which could potentially skew averages upwards. In reality the opposite is probably true, because (1) this deferred compensation that occurs whenever an employee retires is an accurate reflection of what they were earning throughout their career, and so unless a disproportionate number of employees retire and collect payouts in the year under analysis, these payouts belong in the averages, and (2) a significant number of retirees do not work the full year and are therefore screened out based on condition #3 because their “regular pay” did not equal or exceed the “minimum pay for [their] job classification.”
Another potentially distorting factor in these calculations, that, if anything, lowers the averages yielded, is the failure of the three conditions to screen out, for example, City Council members who do not work full time. Most of them have pension contributions and health insurance contributions made by the cities, and their “regular pay” matches or exceeds the “minimum pay for job classification.” But they work part time and their “regular pay” is typically only around $12,000 per year, if that. The presence of these records probably slightly lowers the full-time averages that are calculated.
Because of the sheer size of the pool, even with the weaknesses noted, it is unlikely the results generated are not accurate. They draw from a database that literally includes every single employee under the payroll of any city, county, or state agency in California. In all, 291,011 city employee records were analyzed, 350,150 county employee records, and 239,860 state agency employee records. To verify the methods and the data, the reader is invited to download each of these Excel files, which were created by downloading the State controller’s raw data files and modifying them:
* * *
AVERAGE TOTAL COMPENSATION, FULL-TIME EMPLOYEES
Using this method for several cities in California (ref. Table 1) has validated the accuracy of this method. While a surprising number of employees are excluded from the averages as part-time – usually about half of them – a review of their job descriptions indicates they clearly are not full-time workers: “recreation leaders,” “school crossing guards,” “lifeguards,” “library clerks,” “library pages,” “theater technicians,” “maintenance trainees,” “custodians,” “swim instructors,” “theater arts aides,” and so on.
Table 2, Average Compensation by Entity – 2012
In producing this information, because department classifications are not standardized among cities and counties, it is difficult if not impossible to compile compensation data by type of job. This is possible with individual cities and counties, and yields interesting results. Anyone interested in developing this data for a particular city or county is encouraged to download and study the spreadsheets provided in the CPPC analyses prepared for the following cities.
Download Spreadsheet: Orange_County_Fire Authority_Total_Employee_Cost_2011.xlsx
Discussion/Commentary: The Average Orange County Firefighter’s Total Compensation is $234,000 per Year
Download Spreadsheet: Desert_Hot_Springs_Total_Employee_Cost_2011.xlsx
Discussion/Commentary: Desert Hot Springs, California – Average City Employee Makes $144,329 Per Year
The methods, assumptions, formats and formulas used are the same in all of these analyses; the more recent ones (Palo Alto, Redondo Beach) provide a refined template that is relatively easy to copy and adapt to evaluate any body of Excel data downloaded from the State controller’s website.
* * *
AVERAGE TOTAL COMPENSATION, PUBLIC SAFETY VS. MISC. EMPLOYEES
When evaluating State controller payroll records to isolate full-time employees and develop averages, while it is not practical – because of the size of the databases and the non-standard terminology employed – to develop per-department averages, it is possible to estimate averages for public safety personnel compared to miscellaneous personnel. This can be accomplished by sorting the records to move the full-time records into a single block of data, while doing a secondary sort of the field “pension formula.” It is reasonable to assume that virtually all records showing either “3% at 50” or “3% at 55” pension formulas are for public safety employees. Since very few public safety employees, apart from a still statistically inconsequential pool of new hires in some locales, are on any pension formulas other than “3 at 50” or “3 at 55,” and since very few, if any, employees who are not in public safety are under those pension formulas, this is a reasonably accurate way to separate the records. Doing so yields the results showing on Table 3:
Table 3: Average Compensation, Public Safety vs. Miscellaneous Employees – 2012
* * *
PERSONNEL COSTS AS A PERCENT OF TOTAL BUDGETS
Finally, using the State Controller’s data, one may add all of the employee records, full-time and part-time, to calculate the total personnel expense for all of California’s cities, counties, and state agencies. By comparing the results to data compiled by the California Policy Center in an earlier study “How Big Are California’s State and Local Governments Combined?,” it is possible to estimate what percentage of total government spending is comprised of personnel costs, as shown on Table 4. The amounts reported in the first row of data, “Total Budget,” may be surprising to readers familiar with the numbers, but the basis for them are explained fully in the afore mentioned CPPC study. For example, the direct state budget, once pass-throughs to cities and counties are eliminated, was only $48 billion in 2012. This smaller amount is the appropriate number to use, since the other approximately $50 billion in the state budget are funds that are passed through to cities, counties, school districts and special districts, and are not part of direct state operations.
Table 4: Personnel Costs as a Percent of Total Budgets – 2012 ($=M)
The purpose of this study is primarily to make publicly available a set of compensation averages for California’s state and local government full-time workers. Merely having this benchmark, built from officially reported data, using transparent assumptions, may provide a credible benchmark that can be useful in discussions of what level of pay and benefits is appropriate for California’s public servants.
The data clearly indicates that personnel costs do not, on average, consume 60% to 70% of local budgets, even though in many cities they do consume that much of the budget. On the other hand, at an average equaling 40% of city and 52% of county budgets, personnel costs consume far more than the 8% that has been cited as a reason taxpayers should be unconcerned about public employee compensation. And many cities and counties are now using outside contractors to perform services that are essentially part of normal operations and could be considered personnel costs.
A related observation is that the employer’s share of pension costs, while shown as still consuming less than 10% of total budgets – despite consuming far more than that in many cities and counties – do not include payments on pension obligation bonds. These numbers also don’t reflect how much payments are already scheduled to increase – for example, last year CalPERS announced it would phase in a 50% increase to required pension contributions over the next few years. When one takes this into account, unless all of this increase is borne through increased employee contributions – unlikely – it is necessary to consider the true average compensation for public servants in California’s cities, counties, and state agencies to be at least 10% greater than these estimates.
Finally, total compensation estimates here are skewed downwards because they don’t reflect accruals for the employer’s future retirement healthcare obligations beyond Medicare coverage, which are common for state and local government workers. It is common for these costs to currently range as high as $12,000 per year in retirement, and in most cases they are not pre-funded at all by California’s cities and counties. The present value of these future retirement healthcare obligations constitute an unfunded liability on precisely the same terms as unfunded pension obligations. That is, they represent an accrued cost each year these employees work, which ultimately is translated into cash payments. Appropriate pre-funding of retirement health care obligations probably adds another 3-5% to these estimates of average total compensation.
To reiterate, however, these numbers speak for themselves without requiring embellishment or copious observations. In California, during 2012 the average miscellaneous full-time employee collected total compensation as follows: Cities, $111K; Counties, $98K; State, $90K. Also during 2012, the average full-time public safety employee collected total compensation of: Cities, $170K, Counties, $140K; State, $129K. Add at least 12% to these numbers to reflect unfunded retirement healthcare and pension obligations, and you have an accurate representation of what California’s public servants earn, built from the ground up using the actual payroll records.