is owned and operated by the Economic History Association
with the support of other sponsoring organizations.

Public Sector Pensions in the United States

Lee A. Craig, North Carolina State University


Although employer-provided retirement plans are a relatively recent phenomenon in the private sector, dating from the late nineteenth century, public sector plans go back much further in history. From the Roman Empire to the rise of the early-modern nation state, rulers and legislatures have provided pensions for the workers who administered public programs. Military pensions, in particular, have a long history, and they have often been used as a key element to attract, retain, and motivate military personnel. In the United States, pensions for disabled and retired military personnel predate the signing of the U.S. Constitution.

Like military pensions, pensions for loyal civil servants date back centuries. Prior to the nineteenth century, however, these pensions were typically handed out on a case-by-case basis; except for the military, there were few if any retirement plans or systems with well-defined rules for qualification, contributions, funding, and so forth. Most European countries maintained some type of formal pension system for their public sector workers by the late nineteenth century. Although a few U.S. municipalities offered plans prior to 1900, most public sector workers were not offered pensions until the first decades of the twentieth century. Teachers, firefighters, and police officers were typically the first non-military workers to receive a retirement plan as part of their compensation.

By 1930, pension coverage in the public sector was relatively widespread in the United States, with all federal workers being covered by a pension and an increasing share of state and local employees included in pension plans. In contrast, pension coverage in the private sector during the first three decades of the twentieth century remained very low, perhaps as low as 10 to 12 percent of the labor force (Clark, Craig, and Wilson 2003). Even today, pension coverage is much higher in the public sector than it is in the private sector. Over 90 percent of public sector workers are covered by an employer-provided pension plan, whereas only about half of the private sector work force is covered (Employee Benefit Research Institute 1997).

It should be noted that although today the term “pension” generally refers to cash payments received after the termination of one’s working years, typically in the form of an annuity, historically, a much wider range of retiree benefits, survivor’s annuities, and disability benefits were also referred to as pensions. In the United States, for example, the initial army and navy pension systems were primarily disability plans. However, disability was often liberally defined and included superannuation or the inability to perform regular duties due to infirmities associated with old age. In fact, every disability plan created for U.S. war veterans eventually became an old-age pension plan, and the history of these plans often reflected broader economic and social trends.

Early Military Pensions

Ancient Rome

Military pensions date from antiquity. Almost from its founding, the Roman Republic offered pensions to its successful military personnel; however, these payments, which often took the form of land or special appropriations, were generally ad hoc and typically based on the machinations of influential political cliques. As a result, on more than one occasion, a pension served as little more than a bribe to incite soldiers to serve as the personal troops of the politicians who secured the pension. No small amount of the turmoil accompanying the Republic’s decline can be attributed to this flaw in Roman public finance.

After establishing the Empire, Augustus, who knew a thing or two about the politics and economics of military issues, created a formal pension plan (13 BC): Veteran legionnaires were to receive a pension upon the completion of sixteen years in a legion and four years in the military reserves. This was a true retirement plan designed to reward and mollify veterans returning from Rome’s frontier campaigns. The original Augustan pension suffered from the fact that it was paid from general revenues (and Augustus’ own generous contributions), and in 5 AD (6 AD according to some sources), Augustus established a special fund (aeririum militare) from which retiring soldiers were paid. Although the length of service was also increased from sixteen years on active duty to twenty (and five years in the reserves), the pension system was explicitly funded through a five percent tax on inheritances and a one percent tax on all transactions conducted through auctions — essentially a sales tax. Retiring legionnaires were to receive 3,000 denarii; centurions received considerably larger stipends (Crook 1996). In the first century AD, a lump-sum payment of 3,000 denarii would have represented a substantial amount of money — at least by working class standards. A single denarius equaled roughly a days’ wage for a common laborer; so at an eight percent discount rate (Homer and Sylla 1991), the pension would have yielded an annuity of roughly 66 to 75 percent of a laborer’s annual earnings. Curiously, the basic parameters of the Augustan pension system look much like those of modern public sector pension plans. Although the state pension system perished with Rome, the key features — twenty to twenty-five years of service to quality and a “replacement rate” of 66 to 75 percent — would reemerge more than a thousand years later to become benchmarks for modern public sector plans.

Early-modern Europe

The Roman pension system collapsed, or perhaps withered away is the better term, with Rome itself, and for nearly a thousand years military service throughout Western Civilization was based on personal allegiance within a feudal hierarchy. During the Middle Ages, there were no military pensions strictly comparable to the Roman system, but with the establishment of the nation state came the reemergence of standing armies led by professional soldiers. Like the legions of Imperial Rome, these armies owed their allegiance to a state rather than to a person. The establishment of standardized systems of military pensions followed very shortly thereafter, beginning as early as the sixteenth century in England. During its 1592-93 session, Parliament established “reliefe for Souldiours … [who] adventured their lives and lost their limbs or disabled their bodies” in the service of the Crown (quoted in Clark, Craig, and Wilson 2003, p. 29). Annual pensions were not to exceed ten pounds for “private soldiers,” or twenty pounds for a “lieutenant.” Although one must be cautious in the use of income figures and exchange rates from that era, an annuity of ten pounds would have roughly equaled fifty gold dollars (at subsequent exchange rates), which was the equivalent of per capita income a century or so later, making the pension generous by contemporary standards.

These pensions were nominally disability payments not retirement pensions, though governments often awarded the latter on a case-by-case basis, and by the eighteenth century all of the other early-modern Great Powers — France, Austria, Spain, and Prussia — maintained some type of military pensions for their officer castes. These public pensions were not universally popular. Indeed, they were often viewed as little more than spoils. Samuel Johnson famously described a public pension as “generally understood to mean pay given to a state-hireling for treason to his country” (quoted in Clark, Craig, and Wilson 2003, 29). By the early nineteenth century, Britain, France, Prussia, and Spain all had formal retirement plans for their military personnel. The benchmark for these plans was the British “half-pay” system in which retired, disabled or otherwise unemployed officers received roughly fifty percent of their base pay. This was fairly lucrative compared to the annuities received by their continental counterparts.

Military Pensions in the United States

Prior to the American Revolution, Britain’s American colonies provided pensions to disabled men who were injured defending the colonists and their property from the French, the Spanish, and the natives. During the Revolutionary War the colonies extended this coverage to the members of their militias. Several colonies maintained navies, and they also offered pensions to their naval personnel. Independent of the actions of the colonial legislatures, the Continental Congress established pensions for its army (1776) and naval forces (1775). U.S. military pensions have been continuously provided, in one form or another ever since.

Revolutionary War Era

Although initially these were all strictly disability plans, in order to keep the troops in the field during the crucial months leading up to the Battle of Yorktown (1781), Congress authorized the payment of a life annuity, equal to one-half base pay, to all officers remaining in the service for the duration of the Revolution. It was not long before Congress and the officers in question realized that the national governments’ cash-flow situation and the present value of its future revenues were insufficient to meet this promise. Ultimately, the leaders of the disgruntled officers met at Newburgh, New York and pressed their demands on Congress, and in the spring of 1783, Congress converted the life annuities to a fixed-term payment equal to full pay for five years. Even these more limited obligations were not fully paid to qualifying veterans, and only the direct intervention of George Washington defused a potential coup (Ferguson 1961; Middlekauff 1982). The Treaty of Paris was signed in September of 1783, and the Continental Army was furloughed shortly thereafter. The officers’ pension claims were subsequently met to a degree by special interest-bearing “commutation certificates” — bonds, essentially. It took another eight years before the Constitution and Alexander Hamilton’s financial reforms placed the new federal government in a position to honor these obligations by the issuance of the new (consolidated) federal debt. However, because of the country’s precarious financial situation, between the Revolution and the consolidation of the debt, many embittered officers sold their “commutation” bonds in the secondary market at a steep discount.

In addition to a “regular” army pension plan, every war from the Revolution through the Indian Wars of the late-nineteenth century, saw the creation of a pension plan for the veterans of that particular war. Although every one of those plans was initially a disability plan, they were all eventually converted into an old-age pension plan — though this conversion often took a long time. The Revolutionary War plan became a general retirement plan in 1832 — 49 years after the Treaty of Paris ended the war. At that time every surviving veteran of the Revolutionary War received a pension equal to 100 percent of his base pay at the end of the war. Similarly, it was 56 years after the War of 1812, before survivors of that war were given retirement pensions.

Severance Pay

As for a retirement plan for the “regular” army, there was none until the Civil War; however, soldiers who were discharged after 1800 were given three months’ pay as severance. Officers were initially offered the same severance package as enlisted personnel, but in 1802, officers began receiving one months’ pay for each year of service over three years. Hence an officer with twelve years of service earning, say, $40 a month could, theoretically, convert his severance into an annuity, which at a six percent rate of interest would pay $2.40 a month, or less than $30 a year. This was substantially less than a prime farmhand could expect to earn and a pittance compared to that of, say, a British officer. Prior to the onset of the War of 1812, Congress supplemented these disability and severance packages with a type of retirement pension. Any soldier who enlisted for five years and who was honorably discharged would receive, in addition to his three months’ severance, 160 acres of land from the so-called military reserve. If he was killed in action or died in the service, his widow or heir(s) would receive the same benefit. The reservation price of public land at that time was $2.00 per acre ($1.64 for cash). So, the severance package would have been worth roughly $350, which, annuitized at six percent, would have yielded less than $2.00 a month in perpetuity. This was an ungenerous settlement by almost any standard. Of course in a nation of small farmers, a 160 acres might have represented a good start for a young cash-poor farmhand just out of the army.

The Army Develops a Retirement Plan

The Civil War resulted in a fundamental change in this system. Seeking the power to cull the active list of officers, the Lincoln administration persuaded Congress to pass the first general army retirement law. All officers could apply for retirement after 40 years of service, and a formal retirement board could retire any officer (after 40 years of service) who was deemed incapable of field service. There was a limit put on the number of officers who could be retired in this manner. Congress amended the law several times over the next few decades, with the key changes coming in 1870 and 1882. Taken together, these acts established 30 years as the minimum service requirement, 75 percent of base pay as the standard pension, and age 64 as the mandatory retirement age. This was the basic army pension plan until 1920, when Congress established the “up-or-out” policy in which an officer who was not deemed to be on track for promotion was retired. As such, he was to receive a retirement benefit equal to 2.5 percent multiplied by years of service not to exceed 75 percent of his base pay at the time of retirement. Although the maximum was reduced to 60 percent in 1924, it was subsequently increased back to 75 percent, and the service requirement was reduced to 20 years. As such, this remains the basic plan for military personnel to this day (Hustead and Hustead 2001).

Except for the disability plans that were eventually converted to old-page pensions, prior to 1885 the army retirement plan was only available to commissioned officers; however, in that year Congress created the first systematic retirement plan for enlisted personnel in the U.S. Army. Like the officers’ plan, it permitted retirement upon the completion of 30 years service at 75 percent of base pay. With the subsequent reduction in the minimum service requirement to 20 years, the enlisted plan merged with that for officers.

Naval Pensions

Until after World War I, the army and the navy maintained separate pension plans for their officers. The Continental Navy created a pension plan for its officers and seamen in 1775, even before an army plan was established. In the following year the navy plan was merged with the first army pension plan, and it too was eventually converted to a retirement plan for surviving veterans in 1832. The first disability pension plan for “regular” navy personnel was created in 1799. Officers’ benefits were not to exceed half-pay, while those for seamen and marines were not to exceed $5.00 a month, which was roughly 33 percent of an unskilled seaman’s base pay or 25 percent of that of a hired laborer in the private sector.

Except for the eventual conversion of the war pensions to retirement plans, there was no formal retirement plan for naval personnel until 1855. In that year Congress created a review board composed of five officers from each of the following ranks: captain, commander, and lieutenant. The board was to identify superannuated officers or those generally found to be unfit for service, and at the discretion of the Secretary of the Navy, the officers were to be placed on the reserve list at half-pay subject to the approval of the President. Before the plan had much impact the Civil War intervened, and in 1861 Congress established the essential features of the navy retirement plan, which were to remain in effect throughout the rest of the century. Like the army plan, retirement could occur through one of two ways: Either a retirement board could find the officer incapable of continuing on active duty, or after 40 years of service an officer could apply for retirement. In either case, officers on the retired list remained subject to recall; they were entitled to wear their uniforms; they were subject to the Articles of War and courts-martial; and they received 75 percent of their base pay. However, just as with the army certain constraints on the length of the retired list limited the effectiveness of the act.

In 1899, largely at the urging of then Assistant Secretary of the Navy Theodore Roosevelt, the navy adopted a rather Byzantine scheme for identifying and forcibly retiring officers deemed unfit to continue on active duty. Retirement (or “plucking”) boards were responsible for identifying those to be retired. Officers could avoid the ignominy of forced retirement by volunteering to retire, and there was a ceiling on the number who could be retired by the boards. In addition, all officers retired under this plan were to receive 75 percent of the sea pay of the next rank above that which they held at the time of retirement. (This last feature was amended in 1912, and officers simply received three-fourths of the pay of the rank in which they retired.) During the expansion of the navy leading up to America’s participation in the World War I, the plan was further amended, and in 1915 the president was authorized, with the advice and consent of the Senate, to reinstate any officer involuntarily retired under the 1899 act.

Still, the navy continued to struggle with its superannuated officers. In 1908, Congress finally granted naval officers the right to retire voluntarily at 75 percent of the active-duty pay upon the completion of 30 years of service. In 1916, navy pension rules were again altered, and this time a basic principle – “up or out” (with a pension) – was established, a principle which continues to this day. There were four basic components that differentiated the new navy pension plan from earlier ones. First, promotion to the ranks of rear admiral, captain, and commander were based on the recommendations of a promotion board. Prior to that time, promotions were based solely on seniority. Second, the officers on the active list were to be distributed among the ranks according to percentages that were not to exceed certain limits; thus, there was a limit placed on the number of officers who could be promoted to a certain rank. Third, age limits were placed on officers in each grade. Officers who obtained a certain age in a certain rank were retired with their pay equal to 2.5 percent multiplied by the number of years in service, with the maximum not to exceed 75 percent of their final active-duty pay. For example, a commander who reached age 50 and who had not been selected for promotion to captain, would be placed on the retired list. If he had served 25 years, then he would receive 62.5 percent of his base pay upon retirement. Finally, the act also imposed the same mandatory retirement provision on naval personnel as the 1882 (amended in 1890) act imposed on army personnel, with age 64 being established as the universal age of retirement in the armed forces of the United States.

These plans applied to naval officers only; however, in 1867 Congress authorized the retirement of seamen and marines who had served 20 or more years and who had become infirm as a result of old-age. These veterans would receive one-half their base pay for life. In addition, the act allowed any seaman or marine who had served 10 or more years and subsequently become disabled to apply to the Secretary of the Navy for a “suitable amount of relief” up to one-half base pay from the navy’s pension fund (see below). In 1899, the retirement act of 1885, which covered enlisted army personnel, was extended to enlisted navy personnel, with a few minor differences, which were eliminated in 1907. From that year, all enlisted personnel in both services were entitled to voluntarily retire at 75 percent of their pay and other allowances after 30 years’ of service, subsequently reduced to 20 years.

Funding U.S. Military Pensions

The history of pensions, particularly public sector pensions, cannot be easily separated from the history of pension finance. The creation of a pension plan coincides with the simultaneous creation of pension liabilities, and the parameters of the plan establish the size and the timing of those liabilities. U.S. Army pensions have always been funded on a “pay-as-you-go” basis from the general revenues of the U.S. Treasury. Thus army pensions have always been simply one more liability of the federal government. Despite the occasional accounting gimmick, the general revenues and obligations of the federal government are highly fungible, and so discussing the actuarial properties of the U.S. Army pension plan is like discussing the actuarial properties of the Department of Agriculture or the salaries of F.B.I. agents. However, until well into the twentieth century, this was not the case with navy pensions. They were long paid from a specific fund established separately from the general accounts of the treasury, and thus, their history is quite different from that of the army’s pensions.

From its inception in 1775, the navy’s pension plan for officers and seamen was financed with monies from the sale of captured prizes — enemy ships and those of other states carrying contraband. This funding mechanism meant that the flow of revenues needed to finance the navy’s pension liabilities were very erratic over time, fluctuating with the fortunes of war and peace. To manage these monies, the Continental Congress (and later the U.S. Congress) established the navy pension fund and allowed the trustees of this fund to invest the monies in a wide range of assets, including private equities. The history of the management of this pension fund illustrates many of the problems that can arise when public pension monies are used to purchase private assets. These include the loss of a substantial proportion of its assets on bad investments in private equities, the treasury’s bailout of the fund for these losses, and investment decisions that were influenced by political pressure. In addition there is evidence of gross malfeasance on the part of the agents of the fund, including trading on their on accounts, insider trading, and outright fraud.

Excluding a brief interlude just prior to the Civil War, the navy pension fund had a colorful history, lasting nearly one hundred and fifty years. Between its establishment in 1775 and 1842, it went bankrupt no less than three times, being bailed out by Congress each time. By 1842, there was little opportunity to continue to replenish the fund with fresh prize monies, and Congress, temporarily as it turned out, converted the navy pensions to a pay-as-you-go system, like army pensions. With the onset of the Civil War, the Union Navy’s blockade of Confederate ports created new prize opportunities, and the fund was reestablished, and navy pensions were once again paid from the prize fund. The fund subsequently accumulated an enormous balance. Like the antebellum losses of the fund, its postbellum surplus became something of a political football, and after much acrimonious debate, Congress took much of the fund’s balance and turned it over to the treasury. Still, the remnants of the fund persisted into the 1930s (Clark, Craig, and Wilson 2003).

Federal Civil Service Pensions

Like military pensions, pensions for loyal civil servants date back centuries; however, pension plans are of a more recent vintage, generally dating from the nineteenth century in Europe. In the United States, the federal government did not adopt a universal pension plan for civilian employees until 1920. This is not to say that there were no federal pensions before 1920. Pensions were available for some retiring civil servants, but Congress created them on a case-by-case basis. In the year before the federal pension plan went into effect, for example, there were 1,467 special acts of Congress either granting a new pension (912) or increasing the payments on old pensions (555) (Clark, Craig, and Wilson 2003). This process was as inefficient as it was capricious. Ending this system became a key objective of Congressional reforms.

The movement to create public sector pension plans at the turn of the twentieth century reflected the broader growth of the welfare state, particularly in Europe. As part of the progressive movement, many progressives envisioned the nascent European “cradle-to-grave” programs as the precursor of a better society, one with a new social covenant between the state and its people. Old-age pensions would fill the last step before the grave. Although the ultimate goal of this movement, universal old-age pensions, would not be realized until the creation of the social security system during the Great Depression, the initial objective was to have the government supply old-age security to its own workers. To support the movement in the United States, proponents of universal old-age pensions pointed out that by the early twentieth century, thirty-two countries around the world, including most of the European states and many regimes considered to be reactionary on social issues, had some type of old-age pension for their non-military public employees. If the Russians could humanely treat their superannuated civil servants, the argument went, why couldn’t the United States.

Establishing the Civil Service System

In the United States, the key to the creation of a civil service pension plan was the creation of a civil service. Prior to the late nineteenth century, the vast majority of federal employees were patronage employees — that is they served at the leisure of an elected or appointed official. With the tremendous growth of the number of such employees in the nineteenth century, the costs of the patronage system eventually outweighed the benefits derived from it. For example, over the century as a whole the number of post offices grew from 906 to 44,848; federal revenues grew from $3 million to over $400 million; and non-military employment went from 1,000 to 100,000. Indeed, the federal labor force nearly doubled in the 1870s alone (Johnson and Libecap 1994). The growth rates of these indicators of the size of the public sector are large even when compared to the dramatic fourteen-fold increase in U.S. population between 1800 and 1900. As a result, in 1883 Congress passed the Pendleton Act, which created the federal civil service, and which was passed largely, though not entirely, along party lines. As the party in power, the Republicans saw the conversion of federal employment from patronage to “merit” as an opportunity to gain the lifetime loyalty of an entire cohort of federal workers. In other words, by converting patronage jobs to civil service jobs, the party in power attempted to create lifetime tenure for its patronage workers. Of course, once in their civil service jobs, protected from the harshest effects of the market and the spoils system, federal workers simply did not want to retire — or put another way, many tended to retire on the job — and thus the conversion from patronage to civil service led to an abundance of superannuated federal workers. Thus began the quest for a federal pension plan.

Passage of the Federal Employees Retirement Act

A bill providing pensions for non-military employees of the federal government was introduced in every session of Congress between 1900 and 1920. Representatives of workers’ groups, the executive branch, the United States Civil Service Commission and inquiries conducted by congressional committees all requested or recommended the adoption of retirement plans for civil-service employees. While the political dynamics between these parties was often subtle and complex, the campaigns culminated in the passage of the Federal Employees Retirement Act on May 22, 1920 (Craig 1995). The key features of the original act of 1920 included:

  • All classified civil service employees qualified for a pension after reaching age 70 and rendering at least 15 years of service. Mechanics, letter carriers, and post office clerks were eligible for a pension after reaching age 65, and railway clerks qualified at age 62.
  • The ages at which employees qualified were also mandatory retirement ages. An employee could, however, be retained for two years beyond the mandatory age if his department head and the head of the Civil Service Commission approved.
  • All eligible employees were required to contribute two and one-half percent of their salaries or wages towards the payment of pensions.
  • The pension benefit was determined by the number of years of service. Class A employees were those who had served 30 or more years. Their benefit was 60 percent of their average annual salary during the last ten years of service. The benefits were scaled down through Class F employees (at least 15 years but less than 18 years of service). They received 30 percent of their average annual salary during the last ten years of service.

Although subsequently revised, this plan remains one of the two main civil service pension plans in the United States, and it served as something of a model for many subsequent pension plans in the United States. The other, newer federal plan, established in 1983, is a hybrid. That is, it has a traditional defined benefit component, a defined contribution component, and a Social Security component (Hustead and Hustead 2001).

State and Local Pensions

Decades before the states or the federal government provided civilian workers with a pension plan, several large American cities established plans for at least some of their employees. Until the first decades of the twentieth century, however, these plans were generally limited to three groups of employees: police officers, firefighters, and teachers. New York City established the first such plan for its police officers in 1857. Like the early military plans, the New York City police pension plan was a disability plan until a retirement feature was added in 1878 (Mitchell et al. 2001). Only a few other (primarily large) cities joined New York with a plan before 1900. In contrast, municipal workers in Austria-Hungary, Belgium, France, Germany, the Netherlands, Spain, Sweden, and the United Kingdom were covered by retirement plans by 1910 (Squier 1912).

Despite the relatively late start, the subsequent growth of such plans in the United States was rapid. By 1916, 159 cities had a plan for one or more of these groups of workers, and 21 of those cities included other municipal employees in some type of pension coverage (Monthly Labor Review, 1916). In 1917, 85 percent of cities with 100,000 or more residents paid some form of police pension; as did 66 percent of those with populations between 50,000 and 100,000; and 50 percent of cities with population between 30,000 and 50,000 had some pension liability (James 1921). These figures do not mean that all of these cities had a formal retirement plan. They only indicate that a city had at least $1 of pension liability. This liability could have been from a disability pension, a forced savings plan, or a discretionary pension. Still, by 1928, the Monthly Labor Review (April, 1928) could characterize police and fire plans as “practically universal”. At that time, all cities with populations of over 400,000 had a pension plan for either police officers or firefighters or both. Only one did not have a plan for police officers, and only one did not have a plan for firefighters. Several of those cities also had plans for their other municipal employees, and some cities maintained pension plans for their public school teachers separately from state teachers’ plans, which are reviewed below.

Eventually, some states also began to establish pension plans for state employees; however, initially these plans were primarily limited to teachers. Massachusetts established the first retirement pension plan for general state employees in 1911. The plan required workers to pay up to 5 percent of their salaries to a trust fund. Benefits were payable upon retirement. Workers were eligible to retire at age 60, and retirement was mandatory at age 70. At the time of retirement, the state purchased an annuity equal to twice the accumulated value (with interest) of the employee’s contribution. The calculation of the appropriate interest rate was, in many cases, not straightforward. Sometimes market rates or yields from a portfolio of assets were employed; sometimes a rate was simply established by legislation (see below). The Massachusetts plan initially became something of a model for subsequent public-sector pensions, but it was soon replaced by what became the standard public sector, defined benefit plan, much like the federal plan described above, in which the pension annuity was based on years of service and end-of-career earnings. Curiously, the Massachusetts plan resembled in some respects what have been referred to more recently as cash balance plans — hybrid plans that contain elements of both defined benefit and defined contribution plans.

Relative to the larger municipalities, the states were, in general, quite slow to adopt pension plans for their employees. As late as 1929, only six states had anything like a civil service pension plan for their (non-teacher) employees (Millis and Montgomery 1938). The record shows that pensions for state and local civil servants are for the most part, twentieth-century developments. However, after individual municipalities began adopting plans for their teachers in the early twentieth century, the states moved fairly aggressively in the 1910s and 1920s to create or consolidate plans for their other teachers. By the late 1920s, 21 states had formal retirement plans for their public school teachers (Clark, Craig, and Wilson 2003). On the one hand, this summary of state and local pension plans suggests that of all of the political units in the United States, the states themselves were the slowest to create pension plans for their civil service workers. However, this observation is slightly misleading. In 1930, 40 percent of all state and local employees were schoolteachers, and the 21 states that maintained a plan for their teachers included the most populous states at the time. While public sector pensions at the state and local level were far from universal by the 1920s, they did cover a substantial proportion of public sector workers, and that proportion was growing rapidly in the early decades of the twentieth century.

Funding State and Local Pensions

No discussion of the public sector pension plans would be complete without addressing the way in which the various plans were funded. The term “funded pension” is often used to mean a pension plan that had a specific source of revenues dedicated to pay for the plan’s liabilities. Historically, most public sector pension plans required some contribution from the employees covered by the plan, and in a sense, this contribution “funded” the plan; however, the term “funded” is more often taken to mean that the pension plan receives a stream of public funds from, for example, a specific source, such a share of property tax revenues. In addition, the term “actuarially sound” is often used to describe a pension plan in which the present value of tangible assets roughly equaled the present value of expected liabilities. Whereas one would logically expect an actuarially sound plan to be a funded plan, indeed a “fully funded” plan, a funded plan need not be actuarially sound, because it is possible that the flow of funds was simply too small to sufficiently cover liabilities.

Many early state and local plans were not funded at all; and fewer still were actuarially sound. Of course, in another sense, public sector pension plans are implicitly funded to the extent that they are backed by the coercive powers of the state. Through their monopoly of taxation, financially solvent and militarily successful states will be able to rely on their tax bases to fund their pension liabilities. Although this is exactly how most of the early state and local plans were ultimately financed, this is not what is typically meant by the term “funded plan”. Still, an important part of the history of state and local pensions revolves around exactly what happened to the funds (mostly employee contributions) that were maintained on behalf of the public sector workers.

Although the maintenance and operation of the state and local pension funds varied greatly during this early period, most plans required a contribution from workers, and this contribution was to be deposited in a so-called “annuity fund.” The assets of the fund were to be “invested” in various ways. In some cases the funds were invested “in accordance with the laws of the state governing the investment of savings bank funds.” In others the investments of the fund were to be credited “regular interest”, which was defined as, “the rate determined by the retirement board, and shall be substantially that which is actually earned by the fund of the retirement association.” This “rate” varied from state to state. In Connecticut, for example, it was literally a realized rate – i.e. a market rate. In Massachusetts, it was initially set at 3 percent by the retirement board, but subsequently it became a realized rate, which turned out to be roughly 4 percent in the late 1910s. In Pennsylvania, law set the rate at 4 percent. In addition, all three states created a “pension fund”, which contained the state’s contribution to the workers’ retirement annuity. In Connecticut and Massachusetts, this fund simply consisted of “such amounts as shall be appropriated by the general assembly from time to time.” In other words, the state’s share of the pension was on a “pay-as-you-go” basis. In Pennsylvania, however, the state actually contributed 2.8 percent of a teacher’s salary semi-annually to the state pension fund (Clark, Craig, and Wilson 2003).

By the late 1920s some states were basing their contributions to their teachers’ pension fund on actuarial calculations. The first states to adopt such plans were New Jersey, Ohio, and Vermont (Studenski 1920). What this meant in practice was that the state essentially estimated its expected future liability based on a worker’s experience, age, earnings, life expectancy, and so forth, and then deposited that amount into the pension fund. This was originally referred to as a “scientific” pension plan. These were truly funded and actuarially sound defined benefit plans.

As noted, several of the early plans paid an annuity based on the performance of the pension fund. The return on the fund’s portfolio is important because it would ultimately determine the soundness of the funding scheme and in some case the actual annuity the worker would receive. Even the funded, defined benefit plans based the worker’s and the employer’s contributions on expected earnings on the invested funds. How did these early state and local pension funds manage the assets they held? Several state plans restricted the plans to holding only those assets that could be held by state chartered mutual savings banks. Typically, these banks could hold federal, state, or local government debt. In most states, they could usually hold debt issued by private corporations and occasionally private equities. In the first half of the twentieth century, there were 19 states that chartered mutual savings banks. They were overwhelmingly in the Northeast, Midwest, and Far West — the same regions in which state and local pension plans were most prevalent. However, in most cases the corporate securities were limited to those on a so-called “legal list,” which was supposed to contain only the safest corporate investments. Admission to the legal list was based on a compilation of corporate assets, earnings, dividends, prior default records and so forth. The objective was to provide a list that consisted of the bluest of blue chip corporate securities. In the early decades of the twentieth century, these lists were dominated by railroad and public-utility issues (Hickman 1958). States, such as Massachusetts that did not restrict investments to those held by mutual savings banks, placed similar limits on state pension funds. Massachusetts limited investments to those that could be made in state-established “sinking funds”. Ohio explicitly limited its pension funds to U.S. debt, Ohio state debt, and the debt of any “county, village, city, or school district of the state of Ohio” (Studenski 1920).

Collectively, the objective of these restrictions was risk minimization — though the economics of that choice is not as simple it might appear. Cities and states that invested in their own municipal bonds faced an inherent moral hazard. Specifically, public employees might be forced to contribute a proportion of their earnings to their pension funds. If the city then purchased debt at par from itself for the pension fund when that debt might for various reasons not circulate at par on the open market, then the city could be tempted to go to the pension fund rather than the market for funds. This process would tend to insulate the city from the discipline of the market, which would in turn tend to cause the city to over-invest in activities financed in this way. Thus, the pension funds, actually the workers themselves, would essentially be forced to subsidize other city operations. In practice, the main beneficiaries would have been the contractors whose activities were funded by the workers’ pensions funds. At the time, these would have included largely sewer, water, and road projects. The Chicago police pension fund offers an example of the problem. An audit of the fund in 1912 reported: “It is to be regretted that there are no complete statistical records showing the operation of this fund in the city of Chicago.” As a recent history of pensions noted, “It is hard to imagine that the records were simply misplaced by accident” (Clark, Craig, and Wilson 2003, 213). Thus, like the U.S. Navy pension fund, the agents of these municipal and state funds faced a moral hazard that scholars are still analyzing more than a century later.


Clark, Robert L., Lee A. Craig, and Jack W. Wilson. A History of Public Sector Pensions. Philadelphia: University of Pennsylvania Press, 2003.

Craig, Lee A. “The Political Economy of Public-Private Compensation Differentials: The Case of Federal Pensions.” Journal of Economic History 55 (1995): 304-320.

Crook, J. A. “Augustus: Power, Authority, Achievement.” In The Cambridge Ancient History, edited by Alan K. Bowman, Edward Champlin, and Andrew Lintoff. Cambridge: Cambridge University Press, 1996.

Employee Benefit Research Institute. EBRI Databook on Employee Benefits. Washington, D. C.: EBRI, 1997.

Ferguson, E. James. Power of the Purse: A History of American Public Finance. Chapel Hill, NC: University of North Carolina Press, 1961.

Hustead, Edwin C., and Toni Hustead. “Federal Civilian and Military Retirement Systems.” In Pensions in the Public Sector, edited by Olivia S. Mitchell and Edwin C. Hustead, 66-104. Philadelphia: University of Pennsylvania Press, 2001.

James, Herman G. Local Government in the United States. New York: D. Appleton & Company, 1921.

Johnson, Ronald N., and Gary D. Libecap. The Federal Civil Service System and the Problem of Bureaucracy. Chicago: University of Chicago Press, 1994.

Middlekauff, Robert. The Glorious Cause: The American Revolution, 1763-1789. New York: Oxford University Press, 1982.

Millis, Harry A., and Royal E. Montgomery. Labor’s Risk and Social Insurance. New York: McGraw-Hill, 1938.

Mitchell, Olivia S., David McCarthy, Stanley C. Wisniewski, and Paul Zorn. “Developments in State and Local Pension Plans.” In Pensions in the Public Sector, edited by Olivia S. Mitchell and Edwin C. Hustead. Philadelphia: University of Pennsylvania Press, 2001.

Monthly Labor Review, various issues.

Squier, Lee Welling. Old Age Dependency in the United States. New York: Macmillan, 1912

Studenski, Paul. 1920. Teachers’ Pension Systems in the United States: A Critical and Descriptive Study. New York: D. Appleton and Company, 1920

Citation: Craig, Lee. “Public Sector Pensions in the United States”. EH.Net Encyclopedia, edited by Robert Whaples. March 16, 2003. URL