How Collective Bargaining Affects Government Compensation and Total Spending

Testimony Jobs and Labor

How Collective Bargaining Affects Government Compensation and Total Spending

April 22, 2015 20 min read
James Sherk
James Sherk
Research Fellow, Labor Economics
As research fellow in labor economics at The Heritage Foundation, James Sherk researched ways to promote competition and mobility.

Testimony before Committee on Government Affairs, Nevada Assembly on April 7, 2015

Chairman Ellison and Members of the Government Committee, thank you for inviting me to testify. My name is James Sherk. I am a Research Fellow in Labor Economics at The Heritage Foundation. The views I express in this testimony are my own, and should not be construed as representing any official position of The Heritage Foundation.

Making collective bargaining optional in local governments would benefit Nevadan taxpayers. Local governments face no competitive pressures to operate efficiently. Government unions also have the ability to “elect their own boss” and control both sides of the negotiating table. This dynamic has enabled government unions in Nevada to win more generous benefits than most private-sector workers ever see. Many government employees in the state pay nothing toward their health care or pension benefits. Mandatory collective bargaining unnecessarily inflates state and local spending by $300 per resident. Limiting collective bargaining enabled Wisconsin to close a massive deficit and reduce taxes at the same time. Less far-reaching reforms would still benefit Nevadans.

Collective Bargaining in Government

Nevada law authorizes government unions to act as the “exclusive bargaining agent” of all employees in a local government.[1] This means local governments must negotiate with the union over pay, benefits, and work rules. The locality may not employ workers for anything other than the union-negotiated terms. This gives government unions a monopoly on the labor supplied to local governments. Even if other workers would take the job, the government may not hire them for anything other than union rates. In the event that a government union and a locality cannot agree on a contract the dispute first goes to a fact finder, and then to binding arbitration.

History of Collective Bargaining in Government

This power is a historical anomaly. The labor movement grew out of the difficult working conditions of the industrial revolution. The founders of the labor movement saw unions as a way to prevent corporations from exploiting workers. They also believed that labor and capital were opposed to one another. They saw unions and management as fighting to divide the profits they mutually created. Labor leaders wanted monopoly bargaining power to gain leverage to win more of those profits.

The government operates very differently. Civil service laws mean government employees need not fear exploitive bosses. The government has no profits over which to negotiate. Government unions bargain over how to allocate tax dollars. The arguments for unions in the private sector do not apply to government.

Historically most union leaders concurred with this analysis. Unions rarely attempted to organize the government. In 1919 the American Federation of Labor recognized a few police unions as members of the labor federation. After mass rioting during the Boston Police Strike of 1919 the American Federation of Labor (AFL) revoked the charters of all its member police unions. Through the 1950s most unions had little involvement with government employees. As recently as 1959 only 5 percent of union members worked in government.[2]

Union supporters believed, as AFL-CIO president George Meany declared in 1955, that “[i]t is impossible to bargain collectively with government.”[3] President Franklin D. Roosevelt, who signed the National Labor Relations Act covering private-sector workers, had the same view. In his words, the “process of collective bargaining, as usually understood, cannot be transplanted into the public service.”[4]

In the late 1950s union density began to fall in the private sector. Union leaders quickly came to see organizing government as a potential new source of members and dues. Starting with Wisconsin in 1959 states began giving government unions collective bargaining powers. President Kennedy’s 1961 Executive Order unionizing the federal government spurred many other states to do the same. In the 1960s and 1970s a majority of states granted exclusive representation powers to government unions. Nevada established collective bargaining for local units of government—but not the state government—in 1969. The experience since then has demonstrated the problems government unions can create.

Undermines Representative Government

Collective bargaining in government takes away the final say on public policy from voters’ elected representatives. It forces them to negotiate a contract with union leaders, excluding all other citizens and potential workers from the bargaining table. Consequently, voters’ representatives do not fully control spending and tax decisions. They must reach agreement with union leaders unaccountable to the general public.

For example, voters could elect school board members who campaign on a platform of merit pay for good teachers, meaningful performance evaluations, and trimming employee benefits in order to reduce property taxes. Collective bargaining takes away elected representatives’ power to implement such an agenda. They must obtain the consent of government unions strongly opposed to these policies. If the government and unions disagree, unelected arbitrators have final say over public policy—irrespective of the wishes of the voters’ representatives.

This happened to the Clark County School District (CCSD) in 2012. The school district faced a financial shortfall and had to choose between laying off hundreds of teachers or forgoing pay increases that year. The district’s elected officials wanted to avoid layoffs, but the union insisted on raises. The issue went to arbitration where the arbitrator sided with the union. This forced the CCSD to lay off a thousand teachers.[5]

These layoffs included Edward Savarese, a newly hired fifth-grade teacher at Sewell Elementary School. Savarese’s ability to help underperforming children won him recognition as “New Teacher of the Year.” The District wanted to keep Savarese at Sewell Elementary. However, the union seniority system required CCSD to lay him off instead.[6]

Collective bargaining undermines the principle of voter sovereignty. Elected officials—not unions—should control government operations. Elected officials should decide if teachers like Edward Savarese keep their job or not. Unelected union officers should not make those decisions. Union leaders once believed this, too. As recently as 1959, the AFL-CIO Executive Council stated flatly that “[i]n terms of accepted collective bargaining procedures, government workers have no right beyond the authority to petition Congress—a right available to every citizen.”[7]

No Checks and Balances

Government unions lack the checks and balances that discourage unreasonable contracts in the private sector. First and foremost, unionized businesses must voluntarily persuade their customers to purchase their products. They cannot raise prices too high without losing business to competitors. Private-sector unions demand too much risk bankrupting their employers and eliminating their members’ jobs. Even highly militant unions recognize this reality. The United Auto Workers negotiated famously generous benefits. Nonetheless, the union made deep concessions in 2007 and further concessions during the General Motors and Chrysler bankruptcies to help the automakers stay afloat.

Government unions seldom face this constraint. The government faces little competition for the services it provides. Residents of Reno, Nevada, cannot receive police protection from Carson City or educate their children in Clark County Public Schools. Moreover mandatory taxes fund government operations. The government does not need to hold the price of its operations down to attract customers. Citizens must pay their tax bill or go to jail. This gives government unions considerable power to negotiate inflated contracts. Expensive contracts rarely put their jobs at risk. They simply pass those costs on to taxpayers.

Government unions have a second major advantage over their private-sector counterparts. Private-sector unions cannot choose who they negotiate against. This creates checks and balances at the bargaining table. Unions attempt to maximize employee compensation. Managers try to lower costs and maintain efficient operations. Both sides have strong incentives to reject opposing proposals that go too far.

By contrast, government unions frequently control both sides of the bargaining table. As Victor Gotbaum, head of New York City’s American Federation of State, County, and Municipal Employees (AFSCME) local, explained in the 1970s, “[W]e have the ability, in a sense, to elect our own boss.”[8] Government unions give their political endorsements and campaign support to politicians who give them favorable contracts. They campaign heavily against politicians who try to lower the cost of government services. Consequently, many politicians side with unions during negotiations in order to get re-elected. This dynamic enables unions to win “sweetheart” contracts that private-sector employers would never agree to. These contracts come at the expense of taxpayers.

Inflated Compensation

Collective bargaining has considerably inflated the compensation of Nevada’s local government employees. It has produced benefit packages that few private-sector workers ever see. In many local governments, employees pay nothing toward the cost of their extensive health insurance benefits. For example, the City of Reno pays 100 percent of their employees’ health, dental, and vision premiums.[9] So does the City of Las Vegas.[10] By contrast the average private-sector worker pays over one-sixth of the premiums for a single health plan and over one-quarter of the premiums for a family health insurance plan.[11]

Similarly, government employees in Nevada enjoy retirement benefits that almost no private-sector workers receive. Participants in the Nevada Public Employee Retirement System (PERS) collect pensions equal to 2.5 percent of their average compensation in their highest three years of pay, multiplied by their number of years of service—up to a maximum of 75 percent of their salary. After 30 years they can retire with full benefits. This allows many local government employees to retire in their early fifties with nearly full salaries. The vast majority of private-sector workers whose taxes fund this benefit cannot retire until they become eligible for Social Security at age 66.

In theory government employees are supposed to contribute a portion of their payrolls toward the cost of their pensions. In practice government unions often negotiate to have local governments pay both the employer and employee share of the PERS contribution. For example, the Clark County Public Schools pays both the employer and employee share of retirement contributions.[12]

Even supporters of the union movement recognize the generosity of government compensation in Nevada. Speaking at a University of Richmond Law School panel in 2011 Dr. Jeffrey Keefe—a frequent advocate of collective bargaining in government—explained that his research finds state and local employees make less than comparable private-sector workers, but that Nevada was an exception to that trend:

We again see lower compensation [in state or local government], but it depends, it’s slightly less in the public sector nationally.… I’ve replicated these studies for a variety of states over the last year as well, and they all pretty much show the same thing. There’s a couple of outlier states that I’m still trying to figure out why. Nevada is one of the outlier states.[13]

Other research focusing solely on Nevada’s non-union state employees finds they make 20 percent to 29 percent more than comparable private-sector workers, depending on whether analysts account for the value of greater job security in the government sector.[14] Employees whose unions negotiate to have local government employers cover their share of benefit costs certainly enjoy a greater premium.

Keefe Study Flawed

In recent testimony before the Nevada Assembly Dr. Jeffrey Keefe reversed his position. He now contends that local government employees in Nevada make less than comparable private-sector workers.[15] In new research Keefe argued that local government employees make about 12 percent lower cash wages than comparable private-sector workers, but enjoy more generous benefits. He estimates Nevada’s local government employees receive approximately 6 percent lower total compensation than they would make if they worked in the private sector. However these estimates contain serious flaws.

First, outside researchers could not replicate his results. Andrew Biggs and Jason Richwine—employee compensation experts—used the same data and methodology Dr. Keefe said he used. They found no penalty and perhaps a slight cash pay premium for local government employees in Nevada.[16] This casts serious doubt on the accuracy of his findings.

Second, Dr. Keefe used data from other states to estimate the value of government employee benefits. He imputed benefit cost data from the Employer Costs for Employee Compensation (ECEC) survey for the eight states in the Mountain Census Division to Nevada.[17] From this he drew the conclusion that local governments in Nevada do not overpay their workers. However, other states have different compensation packages for local employees. This approach is especially problematic because many of the states in the Mountain Census Division make collective bargaining optional in local government. For example, in Colorado collective bargaining is permissible but not required for teachers. State law makes no provisions for collective bargaining for any other group of local government employees. Similarly collective bargaining is optional for local governments in Arizona. The cost of government benefits in states with voluntary collective bargaining says little about its cost in Nevada which requires local governments to collectively bargain.

Third, the ECEC survey data that Dr. Keefe relied on ignores a large portion of government compensation. It does not include the value of retiree health benefits. It also does not measure the value of the pensions promised to local government employees. Instead it measures the amount local governments currently pay into retirement systems. The Nevada PERS estimates it only has the resources to pay 71.5 percent of promised obligations.[18] Local governments will have to contribute more in the future to meet their promised obligations. The ECEC ignores these costs.

These mistakes badly flaw Dr. Keefe’s recent estimates. Policymakers in Nevada should not interpret them as an accurate assessment of the cost of government compensation in the state. Most local government employees in Nevada earn considerably more than they would in the private sector.

Higher Spending

Historical experience demonstrates that collective bargaining significantly raises the cost of government for taxpayers. The Heritage Foundation will be releasing a forthcoming Special Report empirically examining the effect of collective bargaining on government spending.[19] My co-authors on the report and I used two methodologies to examine the effects of collective bargaining on government spending. Both approaches showed collective bargaining substantially increases spending.

Estimate of the Effects of Expanding Collective Bargaining on Government Spending in Ohio

The first approach involved case studies of states that substantially expanded collective bargaining in government all at once. For example, Ohio passed comprehensive collective bargaining legislation in 1984. Figure 1 compares the growth of per capita government spending in Ohio to a “synthetic control” of comparable states that did not have similar requirements.[20] As the figure shows, government spending began rising more rapidly in Ohio afterwards—rising to $9,690 in 2012. In the control states, state and local government spending averaged $8,869 per capita in 2012.[21] Ohio taxpayers pay 11.5 percent more for their government than they would without collective bargaining.

More comprehensive evidence comes from regression analysis on panel data. My co-authors used regression analysis to evaluate the effect on collective bargaining powers for government unions on state per capita spending and control variables for the period 1957–2012.[22] This covers the period in which state collective bargaining laws were introduced and became widespread.

The coefficients in the first row of Table 1 shows the effect of a state moving an average of one point up an eight-point scale of collective bargaining powers for government unions. Each point of greater powers is associated with approximately $85 higher per capita state and local government spending for state residents. Making collective bargaining optional for local governments in Nevada would reduce Nevada’s score on this scale by 3.6 points.[23] This implies this policy change would—in the long run—reduce per capita spending by slightly over $300 per resident.

Coefficients: Total State and Local Government Spending

Collective bargaining needlessly inflates costs for Nevada taxpayers. It systematically pressures elected representatives to put the economic interests of government employees ahead of either taxpayers or the recipients of public services. Unsurprisingly, this inflates costs.

Wisconsin’s Experience

Wisconsin Governor Scott Walker eliminated collective bargaining for most state and local employees in 2011. At the time Act 10 passed, these reforms were highly controversial. Union protests captured national attention. However, they became much less controversial after Governor Walker implemented them. His opponent in his successful re-election bid largely avoided the topic during the campaign.[24]

Act 10 became widely accepted because Wisconsinites saw the benefits of the reforms. Governor Walker took office facing a $3.6 billion biennial budget deficit. Act 10 enabled him to close that deficit without tax hikes and without major cutbacks to government services. Instead Governor Walker trimmed the compensation of government employees and reduced state aid to localities by an equivalent amount. Between 2011 and 2013, Act 10 saved state taxpayers $2.7 billion—the vast majority of the spending reductions needed to balance the budget.[25] These savings allowed Governor Walker to pass $2 billion in property and income tax reductions when the economy picked up.

A portion of these savings came from simply operating the government more efficiently. The Wisconsin Education Association (WEA), for example, operates a health insurance company—the WEA Trust. The union used its collective bargaining powers to require districts to obtain health insurance through the trust. Act 10 enabled school districts to shop around for less expensive coverage elsewhere. This produced savings, and pressured the WEA Trust to cut its rates to retain customers. Wisconsin school districts saved tens of millions of dollars on health insurance as a result. For example, after Act 10 the WEA Trust reduced the premiums it charged Appleton Public Schools by 10 percent—with no reduction in teacher benefits.[26]

Nevadan taxpayers would similarly benefit if the legislature loosened unions’ collective bargaining powers. The Clark County Education Association also runs the health benefit trust for Clark County District teachers. The District’s contract requires using this union trust. Making collective bargaining optional would give the District the ability to shop for potentially less expensive coverage from other carriers.[27]


In a democracy government officials—not union officers—should decide public policy. Unions should not have the ability to veto public policies—like performance-based layoffs—they dislike. Unfortunately, mandatory collective bargaining allows them to do just that in Nevada. No matter who the voters elect to local government, they often lack the legal power to implement their agenda.

Making collective bargaining optional would enable local governments to make their own decisions about what best serves the public. This would probably mean far fewer contracts requiring taxpayers to cover 100 percent of employee health premiums or pension payments. It would also mean reducing government spending. State and local spending would fall by $300 per capita if local governments could choose whether or not to collectively bargain.



[1] Nevada Revised Statutes, Chapter 288, Section 160.

[2] Leo Troy and Neil Sheflin, “Appendix A Historical Statistics 1987-1983,” in U.S. Union Sourcebook: Membership, Finances, Structure, Directory, Vol. 1 (West Orange, NJ: IRDIS (Industrial Relations Data and Information Services), 1985).

[3] Leo Kramer, Labor’s Paradox: The American Federation of State, County and Municipal Employees, AFL–CIO(New York: Wiley, 1962), p. 41.

[4] Franklin D. Roosevelt, “Letter on the Resolution of Federation of Federal Employees Against Strikes in Federal Service,” August 16, 1937, April 6, 2015).

[5] Trevon Milliard, “School District: Arbitrator’s Decision Will Lead to Teacher Layoffs,” Las Vegas Review Journal, May 2, 2012, (accessed April 6, 2015).

[6] Trevon Milliard, “Clark County Teacher Honored, then Laid Off,” Las Vegas Review Journal, July 15, 2012, (accessed April 6, 2015).

[7] Kramer, Labor’s Paradox, p. 41.

[8] Ken Auletta, “Who’s to Blame for The Fix We’re In: 4. March 31, 1958: Mayor Robert F. Wagner Issues Executive Order Number 49,” October 27, 1975,,+in+a+sense,+to+elect+our+own+boss.&source=bl&ots=awYLybmhkA&sig=kvhKKg01bq65Px7fS5jtlVHubGs&hl=en&sa=X&ei=_EwcVfiHOKzbsATGiIDgBg&ved=0CC0Q6AEwAg#v=onepage&q=We%20have%20the%20ability%2C%20in%20a%20sense%2C%20to%20elect%20our%20own%20boss.&f=false(accessed April 6, 2015).

[9] City of Reno, Department of Human Resources, “Benefits,” (accessed April 6, 2015).

[10] City of Las Vegas, Department of Human Resources, “Benefits,” (accessed April 6, 2015).

[11] Kaiser Family Foundation, “2014 Employer Health Benefits Survey,” September 10, 2014, (accessed April 6, 2015).

[12] Negotiated Agreement between the Clark County School District and the Clark County Education Association, 2014–2015, Article 29, April 6, 2015).

[13] University of Richmond School of Law, Public Sector Employment in Times of Crisis Conference - Panel 1b, September 27, 2011, minute 13:00 and following, (accessed April 6, 2015).

[14] Andrew Biggs and Jason Richwine, “A State-by-State Ranking of Public Employee Compensation,” American Enterprise Institute Policy Working Paper 2014-04, Tables 6 and 13, April 2014, (accessed April 6, 2014).

[15] Jeffrey Keefe, “Are Nevada Local Government Employees Overpaid?” testimony before the Nevada Assembly of Commerce and Labor Committee, March 25, 2015.

[16] Andrew Biggs and Jason Richwine, “Nevada Local Government Employees Are Not Underpaid,” unpublished memo, April 2015.

[17] These states are Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah, and Wyoming.

[18] Nevada Public Employee Retirement System, “Popular Annual Financial Report: Fiscal Year Ended June 30th, 2014,” p. 8, (accessed April 6, 2015).

[19] This report is authored by Geoffrey Lawrence, Cameron Belt, and James Sherk.

[20] The synthetic control approach involves comparing a group of control states unaffected by a policy change and a treatment state affected by it. The synthetic control algorithm mathematically identifies the best weighted average of states that most closely matches the experience of the treated state. In some cases the algorithm shows that no appropriate control group exists. In the forthcoming Special Report the authors attempted to identify control groups for 10 states that passed comprehensive collective bargaining legislation covering most government employees at the same time: New York (1968), Washington (1968), New Jersey (1969), Oregon (1970), Hawaii (1971), South Dakota (1971), Minnesota (1973), Montana (1974), Florida (1976), and Ohio (1985). Amongst these ten states, we were able to create four synthetic control matches that accurately reflect the traits of the treated units in the pre-treatment period: New Jersey, New York, Ohio, and South Dakota. The results for New Jersey, New York, and Ohio showed positive and significant effects on government spending. The South Dakota results were negative but statistically insignificant. The full results for all states will be published in the forthcoming Special Report.

[21] These control states were Colorado, Indiana, Maryland, Texas, and Wyoming. Note that collective bargaining is permitted but not required for localities under Maryland law.

[22] The full and more detailed results will be published in the forthcoming Special Report.

[23] Nevadan local governments are currently at an “8” on this scale—mandatory collective bargaining with binding arbitration. Making collective bargaining optional but not mandatory would move Nevadan localities to a “3”. State employees are at a “1”—No collective bargaining permitted. The statewide average across state employees, police, firefighters, teachers, and municipal employees would move from 6.2 points to 2.6 points.

[24] Scott Bauer, “Walker’s Challenger Doesn’t Focus on Union Rights,” The Associated Press, April 18, 2014, (accessed April 6, 2015).

[25] Nick Novak, “Update: Act 10 Savings up to $2.7 billion,” The MacIver Institute, October 24, 2013, (accessed April 6, 2015).

[26] Tom Kertscher, “Behind the Rhetoric: The WEA Trust and School Health Care Costs,” PolitiFact Wisconsin, May 21, 2012, (accessed April 6, 2015).

[27] Negotiated Agreement between the Clark County School District and the Clark County Education Association, 2014–2015, p. 46, April 6, 2015).


James Sherk
James Sherk

Research Fellow, Labor Economics