In the US, is the Social Security benefit negatively affected if you have one or a few low income years?

I think it is said that the benefits will be based on the top 35 yearly wages in our work history.

I am not sure if it is by "year" or by "quarter" (meaning 35 x 4 = 140 quarters), but let's say it is by year, it can be the following situation:

1. A person got laid off in February, and didn't find a job until December the same year. So this year, the wage number is really low.

2. Alternatively, it can be: getting laid off in April, and can't find a job for the rest of the year.

3. So we all know if we take the "average" of the numbers, meaning we take the "arithmetic mean" of 35 numbers, it can get pulled down by some low numbers (if the low numbers didn't get eliminated because there are only 32 years of work history, for example, and therefore the low numbers were not excluded from the pool of 35).

4. So as a result, it may turn out that the low number in 2024 or any low numbers, will pull down the calculated average.

5. So in that case, it really is better not to work in 2024, than to work for only a few months.

6. However, it somewhat contradicts the idea that "the more time we work, it should only positively affect our Social Security benefits, not negatively affect it".

7. The situation is similar if it is not "year" but "quarter". If we get laid off in February and get hired again on June 24, then there are two quarters with low numbers: Jan to March, and April to June, with the April to June having a really low number, because there were only 5 working days in that quarter.

So can it negatively affect the benefit number? How does it really work?

So can it negatively affect the benefit number? How does it really work?

No, it cannot. The earnings for past years are already recorded, and if you have 35 years of working history with higher wages than this current year - then it will not affect you at all.

You're confusing smaller rate of growth with loss. The benefits are calculated based on the highest earning 35 years, so if you had 28 years of max earnings and then no earnings at all for 7 years, you'll get less benefits than if you had 28 years of max earnings, then 6 years of no earnings at all, and 1 year of some earnings. So the effect of that low earning year will be smaller than the rest, but still positive.

This is basic math: averaging X over 35 years is always less than (X+Y) over 35 years for any positive value of Y.

• you are saying even if you worked 12 years or 28 years or 32 years, it always is divided by 35? Then if you have only worked for 12 years, it is a really low average because it gets divided by 35. Likewise, if you have only worked for 30 years and it gets divided by 35, then automatically it is like a 14.29% lower average number because you are 5 years less than the 35 years Commented May 21 at 23:34
• @StefanieGauss well of course, if you only worked 12 years you'd get less than if you worked 28 years. Why is it surprising? Commented May 21 at 23:51
• Here's one way to look at it: some of your top 35 years might be zero. So it's the "average of your top 35" not the average of your top non-zero years. Commented May 22 at 0:06
• So it's the "average of your top 35": if it is 0, it will still count, and recorded as 0. If you don't have 35 years of working history, there will be all the 0's to fill in to make it 35 years Commented May 22 at 1:52
• @StefanieGauss that's exactly right. Commented May 22 at 2:10

Your question seems to be based on the premise that if you have fewer than 35 non-zero years, then only those non-zero years will be considered. That is, the average will be calculated by taking a sum, and then dividing by a number that could be less than 35; SS will rank your years from highest earnings to lowest, take the sum until it either reaches a zero-income year or until it reaches the 35 years, and then divide by the number of years that it took.

That is incorrect. SS takes the top 35 years, regardless of whether you made money in all of them. If you make \$70k for ten years and nothing the rest of your life, then they will calculate your average earning as \$20k.

(Although really, all this talk of "averages" is in a sense irrelevant. There is some function f such that your benefits are f(average). This could just as well be expressed as there being some function g such that your benefits are g(total), where g(x) = f(35x). That is, they could just have easily made table with your total earnings as the input and benefits as the output, rather than average as the input, they just figured "You get this much benefits if your average earnings are \$20k" is easier to be immediately meaningful than "You get this much benefits if your total earnings are \$700k" is.)

• yeah, I think the wording of "the average of the top 35 years of top wages" can be somewhat misleading. Some people may take it that if you worked only 10 years, then it is divided by 10. But in fact it is divided by 35 ALWAYS Commented May 22 at 7:10
• "If you make \$70k for ten years and nothing the rest of your life, then they will calculate your average earning as \$20k" --> is more like each year, the income is divided by the median earner. And its the average of best 35 years of that quotient. So \$70K in 2020 will net more benefits that \$70K in 2000. Commented May 23 at 3:08
• @chux-ReinstateMonica huh? the median earner I assume is higher in 2020... so then you'd be dividing by a greater value in 2020, so you'd get a lower benefit from \$70k in 2020? Commented May 23 at 7:28
• @JeopardyTempest Yes it is visa-versa. The key point being that it differs each year: National Average Wage Index (Hmm, I thought it was median wage, looks like the index is average wage) Commented May 23 at 8:34
• @chux-ReinstateMonica I'm actually glad it isn't. If somehow we ever got to a year where 50% + 1 person had no job/earnings (a big enough problem already!)... we'd end up blowing up all the computers with our undefined contributions! :-p Commented May 23 at 9:41