Andrew Biggs has an interesting new report for the American Enterprise Institute investigating the generosity of public-sector pension plans.* He finds that, for the average full-career state worker, traditional defined benefit plans are working quite well and many of these workers are de facto “pension millionaires” because of the amount of money they can expect to receive in retirement.

How does this square with reports from some states that the average pension is quite modest? As we’ve written about for teachers in Illinois and California, the “average” pension is skewed by many employees who qualify for only a very small pension. It’s not accurate to use the statistical average as any indicator of actual payments. To find the typical pension payment, it would be better to look at the statistical median (the midpoint) or even the mode (the most common) amount.

Biggs looks at something different, the payment provided to full-career state government workers. (The piece has an important section reminding readers that the group of workers remaining in the pension plan for a full career is a small group. They are the relative winners in the pension lottery, and all other participants subsidize their pension benefit.) Biggs uses state pension data to look at how well these long-term workers fare under the current system.

There are variations across states, but overall they fare pretty well. In the average state, a full-career state employee receives an annual pension of $36,131. Biggs also converts these figures to total pension wealth and finds that the average full-career state worker can expect to receive $768,940 in pension payments over the course of their retirement.

The numbers are startling, but “generosity” is inherently a judgment call. Another way to look at this question is how much the average pension payment can replace the employee’s pre-retirement earnings. Because retirees tend to face lower expenses than workers, most experts suggest they aim for a 70 percent “replacement rate” in retirement.

The average full-career state employee pension, combined with Social Security where it’s offered, provides an 87 percent replacement rate, meaning these workers can expect to receive $.87 in retirement for every $1 they were earning pre-retirement. This is above the margin commonly suggested by retirement experts and suggests that these workers are “over-saving” for retirement. They’re taking too little of their compensation in the form of present-day salaries and too much in the form of deferred retirement savings. This could be especially problematic for workers with children or those who face other spending constraints, because they’re forced to follow the pension plans’ mandatory contribution rates even if they might prefer more upfront cash and less in savings.

These are two different problems with two different solutions. If the public and policymakers think $36,000 a year in pension payments are too generous, they might support policies that cut retirement benefits and overall compensation for government workers. They may end up with a less generous defined benefit pension plan with all of the same structural problems that exist today. Cutting compensation is also not not likely to help governments attract high-quality workers. That does not mean pensions don’t need to change. Pensions may not be overly generous but they still limit individual choice, fail to provide a secure retirement savings path to all workers, and improperly balance upfront and deferred compensation. But those problems demand different solutions than cutting compensation. For example, policymakers seeking to improve the quality of their workforce would support restructuring retirement benefits while simultaneously raising base salaries. The two positions start in the same place but end very differently. The former would cut overall compensation for government workers, while the latter could be cost-neutral. The end goal should be better, not worse, compensation.

-Chad Aldeman

*To be clear, Biggs focuses only on state government employees, not teachers, although many of the findings would be similar.

This first appeared on



Last updated April 2, 2014