Examining Public Pay in California: The Los Angeles Department of Water and Power

Summary:  The Los Angeles Department of Water and Power (DWP) is the nation’s largest municipal utility, but it may also be one of the clearest examples of excessive public pay driven by powerful public sector unions. This paper analyzes the pay received by DWP employees to their non-DWP counterparts and finds that the average DWP employee receives total compensation that is 155% greater than their non-DWP counterpart.

The largest premiums are found in generic jobs such as custodians, garage attendants, security officers, and the like. The average DWP security officer, for instance, makes 288% more than a non-DWP security officer working in the Los Angeles Metropolitan area. Overall, the weighted average wage premium for DWP employees performing generic jobs was 90% over their counterparts in the Los Angeles area. For all jobs, and including the value of benefits such as pensions and employer paid health insurance costs, the premium for DWP employees as estimated to be 155% higher – that is, 2.5 times as much – than for employees performing work with similar job descriptions in the Los Angeles area.

Applying these premiums to the number of employees at the DWP, the total cost to rate-payers of the DWP paying above market wages is estimated to be $392.8M a year.

*   *   *


The issue of comparing public pay to private pay has challenged academics and sparked fierce debate for years. A serious gap in the academic literature was filled by Andrew Biggs and Jason Richwine’s groundbreaking paper, Overpaid or Underpaid? A State-by-State Ranking of Public-Employee Compensation. Biggs and Richwine found that California State employees receive a total compensation premium of 33% versus their private sector counterpart. Given the scope of their paper, their analysis was limited to state employees only.

However, roughly 90% of all public employees in California work for local agencies. Further, state employees are paid less in wages and receive less generous benefits than local public employees do, suggesting that the bulk of public employees in California receive compensation greater than the 33% premium found at the state level. This paper is the first in a series that will analyze the level of pay for individual public agencies in an attempt to fill this gap.

The Los Angeles Department of Water and Power (DWP) is the nation’s largest municipal utility, serving over four million residents. It is also a powerful example of the above market wages received by California’s public employees. The DWP made national headlines in 2012 and 2013 when Bloomberg reported that their garage attendants were making nearly four times the national average. The Los Angeles Times found that the average total pay for DWP employees was over $100,000 in 2012, approximately 50% higher than other city employees. With recently published 2013 data available on TransparentCalifornia.com, this analysis will update and expand upon the Times’ previous findings.

First, DWP pay is compared to the market in general, as represented by the Bureau of Labor and Statistics (BLS) average wage for the same or similar job, not merely to other government agencies. Secondly, the weighted average of the pay premium found is used to project the total cost associated with the systemic practice of paying above market wages for the department as a whole.

Finally, the value of retirement and health benefits provided to the DWP employee are contrasted to the comparable retirement and health benefits received by a non-DWP employee.

*   *   *


The methodology used is known as the “positions approach” in the economics literature. This approach searches for matching job descriptions and then compares the pay between each. Adjusting for the traits of the underlying people holding the position, also known as the human-capital model, is not utilized. Given the extremely narrow focus of this paper to a single agency in a specific region, as opposed to a state or nation-wide analysis, the “positions approach” is sufficient.

Additionally, the singular focus allows for findings that are based on the actual wages paid, not an estimate based off of a regression analysis.

The DWP employs just over 10,000 people. However, this paper only analyzes employees who worked for a full-year by eliminating any employee with a base salary less than the reported annual salary minimum, leaving 8,318 full-time, year-round employees in 2013.

Twenty-three DWP job titles were selected for analysis, accounting for a total of 3,476 employees. This sample size represents 42% of 2013 full-time, year-round employees and 39% of total payroll expenditures. Job titles were selected by the degree of total employment they represented as well as the ability to reasonably identify a corresponding job title in the BLS report. While identical or similar job title names served as a starting point, the determination in matching a DWP job title to a corresponding BLS title was made entirely on whether or not the job description and responsibilities reasonably corresponded to each.

Modifications were made in the following two cases. First, while the DWP job of customer service representative correlated to the BLS job of the same name, the required skills and job responsibilities for the DWP position appeared much higher than average. Consequently, the corresponding BLS wage was increased to the BLS 75th percentile wage.

A similar adjustment was made for the job of “senior clerk typist.” The DWP job description for both “clerk typist” and “senior clerk typist” correlated to the BLS job of “office clerks, general.” The BLS average wage was used as the comparison for the DWP job of “clerk typist” and the 90th percentile wage was used for the comparison against the DWP’s “senior clerk typist” position.

An appendix listing the exact comparisons made is included at the end of the paper. The analysis was able to incorporate the seven most populated job titles held within the DWP, along with 16 additional job titles of various sizes.

*   *   *


Regular Pay

The DWP wage is compared to the average wage for the same or comparable job as reported in the May 2013 wage estimates by the Bureau of Labor and Statistics for the Los Angeles Metropolitan area.

The BLS wage does not include overtime pay, but does include a variety of additional pays such as longevity, hazard, and incentive pay. As such, only the DWP wages without overtime pay (Regular Pay) should be considered as analogous to the BLS wage. Regular Pay is defined as base pay plus the multitude forms of routine “other pay” that DWP employees receive. In 2013, the average non-OT earnings (Regular Pay) of a DWP employee were $99,900.

 Table 1  –  Average DWP Regular Pay vs Average BLS Wage by Job Title


Overtime Pay

Overtime pay was excluded from this analysis to create parity between the DWP wage and BLS estimates. However, the DWP provides overtime pay at a higher rate than even firefighters or police, which casts serious doubts about the management structure and the necessity of the overtime pay issued. Including overtime pay increases the average 2013 total earnings of a DWP employee by 15% – to $114,941.

Additionally, an incomprehensible 92% of DWP employees receive overtime pay of some kind. The Los AngelesTimes discovered at least one particular example which confirms that at least some of the overtime pay is the result of union-friendly contracts, not necessity, when they revealed that DWP employees receive overtime pay for work that an outside contractor performs. This bears repeating: DWP employees can receive overtime pay for work that others do. Such a provision is unheard of in typical labor contracts, according to the expert cited in their article.

Despite the fact that the overtime pay at the DWP is at least partially excessive and not reflective of a genuine staffing need, it is impossible to quantify the proportion that is driven by abuse, as compared to that which is driven by legitimate need. Consequently, overtime pay is omitted from this comparison. Still, it must be noted that the average DWP employee receives a non-trivial benefit, averaging 15% in 2013, from the department’s atypically generous overtime policy.

A 90% pay premium for regular jobs

Many of the jobs with the smallest degree of DWP premiums are likely due to the fact that the DWP essentially is the market for that position in the Los Angeles Metropolitan area. Electric distribution and electrical mechanics and their comparable BLS job titles (listed in the Appendix) are jobs unique to a utility company. Obviously the DWP is, by far, the largest utility in the Los Angeles Metropolitan area. Consequently, the BLS comparable wages are going to be overwhelmingly represented by DWP employees, making a comparison less meaningful.

To have a more accurate picture of the above market wages paid by the DWP it is necessary analyze generic jobs that have a robust, non-DWP market such as custodians, security officers, and the like. When filtering for jobs not unique to a utility company, the Regular Pay received by a DWP employee is 90% greater than the market average.

Table 2  –  Average DWP Regular Pay vs BLS Wage for Generic Jobs Only



It is well documented that public sector defined benefit plans provide more generous benefits than a defined contribution plan; typically the public employee shares in the cost of funding this plan to some extent. In CalPERS, for instance, many public employees pay half of the required contribution rate, which can range from 10-30% of salary, depending on the individual employing agency.

DWP employees, however, participate in their own pension plan and contribute only a maximum of 6% of salary towards their defined benefit plan, with earlier employees paying even less than that. While the DWP plan offers benefits slightly lower than CalPERS in absolute terms, the ability to receive these benefits at a reduced cost to the employee greatly increases the net value of the DWP pension plan.

However, most employees, particularly private employees, participate in a defined contribution plan, which is vastly less generous than the defined benefits plans California’s public employees participate in. A comparison to the type of benefits typically received from a contribution of 6% of salary is illustrative in that regard.

Given private employees must contribute 6.2% of their salary in Social Security taxes, any 401(k) style matching retirement benefits available to the non-DWP employee are only available if they incur an additional cost. Consequently, the following will compare the retirement benefits available based on the assumption that each employee is contributing the same level of salary (6%) towards their retirement.

Take, for instance, the average DWP custodian receiving an average base pay of $52,734 vs. the BLS average wage of $26,810. Many custodial positions outside of the DWP are hourly and do not offer benefits of any kind. However, the DWP custodian receives employer-paid medical benefits and is enrolled in a “2.3% @ 55” defined benefit plan. Assuming the DWP custodian retires at the age of 60 with 30 years of service, they will begin receiving an annual pension of $36,386. By contrast, the private sector custodian will not be able to receive anything from Social Security until the age of 62, at which point they will receive a yearly benefit of $8,880 based on a wage of $26,810.

Using an average life expectancy of 85, a discount rate of 3.75%, and an annual Cost of Living Adjustment (COLA) of 2%, the net present value of the DWP custodian’s pension benefit is $706,841. The net present value of the non-DWP custodian’s Social Security benefit is $161,250. This represents a retirement benefit for the DWP employee nearly 340% greater than that of what their non-DWP counterpart can expect to receive, despite having contributed the same percentage of salary towards their retirement plan.

The wages used to compute the DWP’s pension are base salary only and do not include the various forms of “other pay” that are included in Regular Pay for the salary comparisons done above.

The table below displays the results of the same analysis for five of the most populated job titles in the DWP.

Table 3 – Net Present Value of Avg Full-Career DWP
Retirement Benefit vs Comparable Social Security Benefit


While it may seem initially counterintuitive that the largest pension premium is not found in the position with the largest wage premium, it makes sense when you remember there is a cap on the maximum Social Security benefit. As such, the key driver is the absolute value of the wages used for the DWP employee.

The comparison in Table 3 assumes an employee had worked 30 years and retired at the average wage reported, which is unlikely. Granted, this weakness applies to both sides equally in that both the DWP and non-DWP employee would likely retire at a higher wage than the average. Still, it is sufficient to illustrate the enormous disparity in pension benefits offered to DWP employees in relation to the cost of an annual contribution of no more than 6% of salary.

Total Compensation Premium

An alternative and more robust measure of comparison would be to calculate the value of employer-paid benefits as a percentage of annual wages.

To do so, we rely on the model pioneered by one of the nation’s leading experts on public sector pay and pensions, Andrew Biggs. An explanation as to how we calculated the value of the DWP’s defined pension benefits can be found starting on page 26, with an explanation on how we treated the value of the Social Security benefit beginning on page 40. The most pertinent section is reprinted below:

“To calculate pension compensation paid from state government pensions, we must convert normal costs as published by those plans to a measure using risk-appropriate discount rates. To do so, we gathered data on over 20 plans from California, Florida, Colorado, Washington, and Rhode Island in which pensions’ own actuaries have calculated pension costs under different discount rates. The median result indicates that a 1 percentage point reduction in the discount rate raises the normal cost of a plan by around 36 percent. As a check, we performed our own calculations using workers stylized to be typical of state government employees, which produced similar results.

The factor to convert a normal cost would equal 1.36(re – rra), where re equals the expected return on plan assets and rra the risk-adjusted discount rate. For instance, the factor to convert a normal cost calculated at 8 percent to a 4 percent discount rate would be 1.364, = 3.42. From this risk-adjusted total normal cost we subtract the value of employee contributions to arrive at net pension compensation. For instance, a plan with a total normal cost of 10 percent of wages at an 8 percent discount rate would have a normal cost of 34.2 percent of pay using a 4 percent discount rate. If the employee contributes 5 percent of pay to the plan, his net pension compensation would be equal to 29.2 percent of wages.”

In a nutshell, public pension systems understate the true cost, and value, of their benefits by using an inappropriately high discount rate in their actuarial calculations. Experts from the Congressional Budget Office, Federal Reserve Board, federal Bureau of Economic Analysis, Moody’s Investment Services, and across academia agree that an appropriate discount rate is one that matches the risk characteristics of the benefit (sidebar 1.)

The most commonly used discount rate for benefits that have little to no risk is the yield on a 20 year Treasury. Currently, the 20 year Treasury rate is at historic lows in the low 2 percent range. Therefore, we will again follow Biggs’ lead and use 4 percent as our discount rate, roughly the average yield over the past decade of a 20-year Treasury.

The DWP pension plan has a normal cost of 23.85% of wages (page 12.) The plan also uses a discount rate of 7.5% for its actuarial calculations (page 40.)  However, given the DWP’s pension benefits are guaranteed, and thus have no risk, it is necessary to adjust this cost to a 4 percent discount rate for the reasons outlined above. This represents a discount rate reduction of 3.5%. Per Biggs, every 1 percent point reduction increases the normal cost of a plan by around 36%. Therefore, the factor to convert to a risk-adjusted total normal cost would be 1.363.5,= 2.93. Multiplying 2.93 by the normal cost of 23.85% = 69.96%. After subtracting the contribution rate for DWP employees of 6%, the net pension compensation is worth 63.96% of wages.

For health benefits, the DWP’s 2013 payroll report stated the average cost per employee was $16,230. Therefore the total compensation for the DWP employee is: (Regular Pay * 1.6396 + $16,320.)

For the BLS counterpart, data from the BLS Employer Costs for Employee Compensation (ECEC) Survey was used to estimate the cost of employer-paid health insurance at 11% of wages. The total cost of Social Security and employer contributions towards a matching 401(k) was estimated at 9% of wages, also using data from the ECEC Survey. The formula for BLS Total Compensation is: (BLS Wage * 1.2.)

The chart below is a graphical representation of the total compensation premium for 10 of the most heavily populated positions in the DWP. Table 4 documents the total compensation premium found for all job titles analyzed in this paper. The average DWP employee receives compensation that is 155% greater than their non-DWP counterpart.



Table 4  –  Total Compensation for DWP Employee vs. BLS Counterpart


*   *   *


This paper is limited in that no attempt to compare the value of retiree health benefits, fringe benefits, or job security is made. However, it is extremely likely that doing so would only inflate the compensation premium already found.

These benefits, particularly things like job security, a generous (unlimited until mid-2013) sick day policy, and so forth are all extremely likely to weigh in favor of the DWP employee.

Another limitation was that many jobs in the DWP do not have a meaningful non-DWP counterpart in the Los Angeles area. Most notably, the jobs of “Water Utility Worker” and “Electric Station Operator” were omitted, despite being the eighth and ninth most populated jobs in the DWP, given the lack of a suitable BLS counterpart.

*   *   *


In order to project the total cost estimated with the DWP’s routine policy of paying above market wages, it is necessary to look at the actual dollar cost found as a percentage of total payroll.

For instance, the actual total cost of the pay premium received by the 3,476 DWP employees analyzed in this paper was: $116,668,950. This number was created by multiplying the difference between the BLS wage and the DWP Regular Pay by the total number of employees for each job title, as displayed in Table 1.  After which, one can simply sum the actual cost found for each of the 23 job titles for a total cost.

The total non-OT payroll for these employees is $323.8 million. As such, the total pay premium found represents 36.02% of payroll. If we assume these findings hold true for the rest of the DWP, the total cost can be estimated by multiplying the total (including PT employees) non-OT payroll of $927.6 million by 36.02% for a total cost of $334.1 million.

Further, the savings associated with reducing pay to the market average would expand beyond the immediate reduction in payroll expenditures; as doing so would simultaneously reduce the cost of pension benefits.

The DWP estimates their total normal cost for 2014 to be 17.56% of payroll (page 12.) As a result, every dollar reduction of payroll would save the DWP an extra 17.56 cents in pension contributions.

Assuming the DWP was to pay market wages, the total savings would rise to $392.8M when including the savings from the reduced pension contributions.

New Salary Contract

The public pressure from the Los Angeles Times’ exhaustive work in exposing the DWP’s unlimited sick day policy, overtime pay abuse, and high levels of pay led to a new contract being signed in late 2013.

Revealingly, the “concession” by the DWP only further demonstrates the power of their union. The deal slightly reduced benefits for new hires only; salaries were not reduced at all. Current employees received no benefits or pay reduction of any kind and merely saw their yearly base salary raise delayed until 2016.

As mentioned above, the DWP employee receives over 600 forms of other pay beyond their base salary, not to mention overtime pay that accounted for an average 15% increase in total earnings in 2013. The deal did not touch any of these forms of pay.

*   *   *


The above analysis makes clear the DWP’s primary motivation in setting employee compensation is political in nature – specifically, to accommodate the union and its members at the expense of ratepayers.

Particularly alarming is that the DWP is funded via what amounts to a regressive tax by being able to set the cost of water and power for their over 4 million customers. Consequently, Los Angeles area residents are being forced to subsidize lavish compensation packages that dwarf their own incomes.

The theoretical case against public sector unions is strong, with many of the Left’s greatest heroes having previously warned of the danger and impossibility of collectively bargaining with the public at large. The DWP offers a specific, real-world example of just how egregiously public sector unions can enrich themselves at the expense of the public.

*   *   *

About the Author:  Robert Fellner is Research Director for TransparentCalifornia.com, a joint project of the California Policy Center and the Nevada Policy Research Institute.

*   *   *


Please click here to go to the BLS’ Los Angeles Metropolitan Area May 2013 overview page, which contains the wage estimates listed below. Clicking the link for the individual job titles provides a description of the job’s duties and responsibilities, but wage data on a national level only.

DWP Job TitleBLS WageBLS Job Title
Senior Clerk Typist$45,960Office Clerks, General (90th percentile)
Customer Service Representative$44,800Customer Service Representatives (75th percentile)
Electric Distribution Mechanic$97,210Electrical Power-Line Installers and Repairers
Electrical Engineering Associate$61,530Electrical and Electronics Engineering Technicians
Electrical Mechanic$88,680Electrical and Electronics Repairers, Powerhouse, Substation, and Relay
Civil Engineering Associate$62,080Civil Engineering Technicians
Security Officer$26,640Security Officers
Meter Reader$46,930Meter Readers, Utilities
Custodian$26,810Janitors and Cleaners, Except Maids and Housekeeping Cleaners
Heavy Duty Truck Operator$42,210Heavy and Tractor-Trailer Truck Drivers
Clerk Typist$31,350Office Clerks, General
Electrical Engineer$110,620Electrical Engineers
Secretary Legal$54,010Legal Secretaries
Plumber$65,350Plumbers, Pipefitters, and Steamfitters
Air Conditioning Mechanic$54,690Heating, Air Conditioning, and Refrigeration Mechanics and Installers
Truck Operator$42,210Heavy and Tractor-Trailer Truck Drivers
Civil Engineer$95,250Civil Engineers
Custodian Supervisor$42,210First-Line Supervisors of Housekeeping and Janitorial Workers
Garage Attendant$29,175Approximated this value*
Mechanical Engineer$98,330Mechanical Engineers
Management Analyst$90,040Management Analysts
Maintenance and Construction Helper$44,930Helpers, Construction Trades, All Others
Mechanical Engineering Associate$62,520Mechanical Engineering Technicians

*The BLS reported a mean wage of $21,320 for a parking lot attendant. However, the Garage Attendant for the DWP entails additional non-skilled auto mechanic responsibilities and the wage was increased to reflect that by averaging the wage for a parking attendant and automotive service mechanic.

2 replies
  1. Josh
    Josh says:

    the average security officer for private is not armed. The average security officer for a public sector position is armed. Armed guards are paid a premium over unarmed guards that does not seem reflected in the adjustments in your report.

    • Robert Fellner (@GrimHogun)
      Robert Fellner (@GrimHogun) says:

      Hi Josh,

      I compared the job description for each. I also reviewed the job requirements and prerequisites so if the DWP had something like a requirement to have a valid firearms license, training in firearms, etc. I would have certainly made an adjustment or excluded them altogether.

      Here is DWP description:

      A Security Officer patrols and safeguards City-owned buildings, parks, reservoirs, facilities, grounds, and equipment; answers questions and directs visitors/patrons; assures that only authorized persons enter buildings, yards, and other restricted areas; prepares reports and keeps records; may make preliminary investigations of accidents, thefts, and disturbances; may issue warnings or citations for illegal parking; may testify in court and may detain violators pending the arrival of a peace officer.

      Which is very similar to BLS description: Guard, patrol, or monitor premises to prevent theft, violence, or infractions of rules. May operate x-ray and metal detector equipment. Excludes “Transportation Security Screeners” (33-9093).

      You can click here to see the rest, there is no mention of a firearms license or anything like that:


      Also here is an article stating that the DWP security officers “do not carry firearms and are not sworn police…” http://www.naylornetwork.com/app-ppd/articles/index-v2.asp?aid=175105&issueID=23092

      Also, in an older article from 2004 it was found that:

      “They very emphatically stated that almost anyone can get a job as a security officer within [DWP]…” the internal report said… the DWP is reevaluating the use of contract guards and is looking at whether to upgrade the status of security officers, who are city employees, to “public safety officers,” a state designation that requires more extensive training and rigorous background checks.” http://articles.latimes.com/2004/oct/13/local/me-dwp13

      I mention that only to suggest that if the DWP’s security officers were more trained, carried a firearm, etc. as compared to regular security officer, it sounds like they would be classified as a “public safety officer” not just security officer.

      Let me know if you see anything else that needs a closer look.


Trackbacks & Pingbacks

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *