The best answer to the question “how does the federal employee retirement system work?” soon may be the same as the answer to the question “how long is a piece of string?”

It depends.

Now, that’s not how you go about things if you were starting an organization from scratch and counting on it to provide vital services. You would design a straightforward retirement program in which employees put in a certain amount of their salary, the employer did the same, and the eventual benefit would involve a straight-forward calculation based on length of service and salary. Maybe for occupations where careers are shorter because the work is especially physically demanding or dangerous you would tweak the required contribution and the payout formula to take that into account.

That way everyone would understand the terms of the deal when they decided whether to take the job and how long to stay in it—to leave at some point or to stick around all the way until retirement. Everybody on the same page, no potential for dissention and resentment among your workers over their benefits. They could just concentrate on going about their work knowing that they all are being treated fairly.

Oh, wait. That’s what the government used to have—had for six decades. Now, it depends.

First, it depends on whether you first joined the government before 1984. If you did, you are now a member of minority, less than a tenth of the workforce and shrinking fast, as someone in the Civil Service Retirement System. Unless you switched to the Federal Employees Retirement System voluntarily at some point. Or unless you had a break in service under certain conditions and returned as CSRS-Offset.

The point is, things started changing when Congress and the Reagan administration decided to shore up Social Security by requiring more people to pay into it. For those hired in 1984 and after, that included federal employees.

FERS and CSRS have many differences beyond Social Security coverage. Just to name a couple of the most importance, most FERS retirees only receive a COLA after they reach age 62 (with certain exceptions) and when the COLA is 3 percent or higher, FERS retirees receive the COLA minus 1 percent. If it’s between 2 and 3 percent, FERS retirees only receive 2 percent. If it’s less than 2 percent.

FERS was designed to produce a benefit about equal to that of CSRS, counting Social Security and the value of employer contributions toward the TSP, which CSRS people don’t receive. To get those roughly equal benefits, FERS was designed to require the same total that CSRS people pay. For CSRS, 7 percent, all going into the civil service retirement fund, for FERS, 0.8 percent into that fund and the standard 6.2 going into Social Security.

That arrangement managed to stay stable for nearly three decades until Congress and the Obama administration decided to increase the required contribution for FERS employees in the name of deficit reduction (the more employees pay in, the less the government pays in as the employer). There were two separate increases, in both cases applying only to those hired starting the following year.

For those hired after 1983 but before 2013, that contribution is still 0.8 percent; for those hired in 2013 it is 3.1 percent; for those hired after 2013 it is 4.4 percent. (Although in the latter two cases, the required contribution is only 0.8 percent if they were rehired after a break in service and had at least five years of prior creditable civilian federal service. Got that?)

All of this for the same level of benefit per length of service and high-3 salary, mind you.

Now, proposals that used to be relegated to the far back benches of Congress are moving forward, to eliminate the defined benefit portion entirely. The reason, of course, is deficit reduction—removing a future obligation of the government to pay the benefits.

The idea is to follow the private sector’s unfortunate example of racing to the bottom by providing only a defined contribution benefit—the TSP being the equivalent of 401(k)s that have largely replaced traditional annuities.

As part of this tradeoff, the government contribution toward those new employees’ benefits might be increased—might is the operative word; backers of this idea say they still are mulling the numbers.

With such an increase and with diligent and successful personal investing, those newly hired employees might achieve retirement savings that would produce an equivalent to the FERS defined benefit, on converting those savings to an annuity at retirement—might again being the operative word.

What this would mean to the government’s ability to attract new employees in competitive fields such as research, IT, cybersecurity and others remains to be seen. If the switch from CSRS to FERS—a much less draconian change—is a guide, the reaction will be strongly negative and unpleasant, with the younger employees resenting the older ones and some of the older ones unfortunately making things worse by gloating, subtly or not. Even the boost in the amount of required contributions in FERS depending on hiring date produced that reaction.

Dress it up however you might—as the opportunity to better manage your own retirement savings, as a crucial national need in times of high deficits, whatever—people recognize when they’re getting the short end of the string.