The election of Donald Trump, coupled with Republican majorities in the House and Senate, means that there is a great deal to report and analyze in the domestic news cycle. But the outgoing government is not done yet, so the stage is set for some very important things to sneak under the political radar.

One such thing is HR 82, the Social Security Fairness Act, which passed the House by an overwhelming margin (327-75) on November 12. This bill contains a great deal of fine print, but one provision should be more than sufficient to produce opposition by anyone who claims to care about sustaining the long-term health of Social Security. It also demonstrates how the lack of a clear vision for the Social Security Program has led to ad hoc changes to the program that are long on political gaming and short on fiscal responsibility.

Section 3 of the Social Security Fairness Act repeals the Windfall Elimination Provision (WEP) of the 1983 Social Security Amendments to the original law created with FDR’s signature in August of 1935. The Social Security program got into trouble in the early 1980s and a number of changes were instituted to stabilize the program. The most well-known change was to phase in an increase in the full-benefit retirement age from 65 to 67, to adjust the program for the fact that Americans were living longer.

One of the lesser-known changes made in 1983 was the WEP. It addressed a bias in the procedure for calculating a given person’s monthly benefit that led to an unfair outcome while also creating additional stress on the budget.

Most state and local government employees have generous pensions provided either by their government employer or their labor union. Workers covered by such pension programs were generally not required to participate in the Social Security program, which was fine with these workers since this meant they would not have to pay Social Security payroll taxes. It was also fine with unions that managed these pensions because money paid into the Social Security program was not under their control, while money paid into their pension plan was.

If all workers worked in covered employment, and therefore paid into Social Security over their earnings lives, or if all workers worked in uncovered employment and therefore did not pay into Social Security but got a generous pension instead, none of this would be problematic. But a common situation was that a person would work for years without paying any Social Security payroll taxes because he or she would be getting a generous pension, but would then retire and go to work in another capacity that was covered by Social Security. If they worked for more than 10 years in such covered employment, they would qualify for a Social Security pension, too.

This is where things get a little tricky. Social Security computes a monthly benefit level based on prior earnings. This begins with adjusting every year’s earnings upon which Social Security payroll taxes were imposed to account for the effects of inflation to express all of these amounts in terms of the purchasing power of the year the individual turns 60 (someone who made $150,000 per year in 1980 made much more money in real terms than someone who made $160,000 in 2010). Social Security then selects the 35 highest-valued years and converts this into a monthly average by dividing by 420 (35 years times 12 months per year). Social Security then applies a generous replacement rate of 90 percent for the first X dollars of this number, 32 percent for the next Y dollars, and 15 percent for any remaining monthly income.

The X and Y numbers are called bend points. Because they account for the effects of inflation, they are different for every year. These replacement rates function like progressive income tax rates, because the idea is that those who have the least money need the most generous treatment (at 90 percent benefit, these workers get nearly all they were earning replaced in retirement) and those who have the most money need the least generous treatment (at 15 percent, these workers get comparatively little of what they were earning before in retirement).

The problem with this procedure is that when people have uncovered income for which they paid no Social Security payroll taxes, and also have 10 or more years of covered income for which they did, many of their top 35 years will be filled with zeros because the calculation of the average monthly income only accounts for covered income (income earned that was subject to the Social Security payroll tax).

This means that a person who has a very large government pension over employment years in which they paid no Social Security payroll taxes, and also has 10 or more years over which did, they appear relatively poor because of all of those zeros in the 35 years in the benefit calculation. Since the program replaces more of their income if they are poor (90 versus 32 or 15 percent), people with a large pension paid for with uncovered employment might also get the most generous terms on their additional Social Security pension.

To see why this is unfair, imagine you and I are exactly the same age, and that we started our second job at the same company at the same time doing the same thing for the same pay. Suppose we then retire on the same day and start collecting Social Security immediately. Without some kind of adjustment, our Social Security checks would be identical, even if the pension I now get for my first job is huge and yours is tiny or non-existent.

If the replacement rate were, say, 75 percent for everyone and all income, perhaps that would be fair. But that’s not the case. Since the most generous tier (90 percent) covers the first X dollars of average income, retired government employees can collect pension dollars for the years they didn’t pay into Social Security, while receiving payouts as if they had. These double-dippers could keep more total income than workers who paid in over their whole lives, and likely earned much less. Social Security was paying the relatively rich at a very generous rate, and in so doing costing the budget dearly.

In 1983, the WEP tried to address this unfairness by discounting the replacement rates for double-dippers. That fix was ad hoc in nature and its implementation was (and is) complicated, but suffice it to say that since 1983 much of the advantage enjoyed by double-dippers due to their being three different replacement rates is now removed. This means that, after we adjust for inflation, double-dippers today get lower Social Security monthly checks than their identical twins who retired in, say, 1975.

Paying less to double-dippers saved the Social Security Program’s budget a great deal of money. As recently reported in The Wall Street Journal, the Congressional Budget Office estimates that over the next 10 years the cost of eliminating the WEP will cost the Social Security budget $196 billion. To put this in perspective, this is more than double what could be saved by increasing the full benefit retirement age from 67 to 70.

As often happens this time of year, especially right after elections, foolish bills are introduced with too little time for proper consideration. The more complex the program involved, the easier it is to sneak something by.

While the WEP is far from perfect, it at least pushes back against a bias introduced by having monthly income replacement rates that are aimed to redistribute income from the relatively wealthy to the relatively poor.

There is profound irony in all of this. If we do nothing, the Social Security Administration estimates that benefits will have to be trimmed by about 20 percent across the board in the future (the earlier they are reduced, the less they need to be reduced). Those who are poorest will be harmed most by this outcome.

With an imminent vote in the Senate and an outgoing president who will surely sign the Social Security Fairness Act into law, public sector pension employees who double-dip will be the only ones helped by removing the WEP. This will worsen the already insolvent Social Security budget and require that benefits be trimmed even sooner than otherwise and/or by a greater amount than otherwise. This will harm those at the bottom the most. It will effectively transfer wealth from the poorest recipients of Social Security benefits to the richest, hardly what supporters of Social Security claim to favor.

No matter how one feels about Donald Trump, it is clear that many who voted for him did so in part because they have grown tired of this kind of last-minute budget game playing. Too often, as now, it benefits a relatively small interest group (fewer than 3 million public-sector double-dippers will receive a windfall from the rule change) at the expense of everyone else, while worsening our budget problems.