I think you are overplaying the whole widespread suffering and human misery angle. In 1934 when SS was passed the suffering and misery was a direct result of the Great Depression. Before then the elderly either continued to work or lived with relatives or friends. The GD made finding work for the elderly much more difficult and made it difficult for relatives to take care of older loved ones. Once the GD passed SS would have been entirely unnecessary again for the most part. As with any government entitlement however, once put into place it can't be taken back.
The 1930's census showed that 58% of men over 65 still worked. By 2002 that figure dropped to 18%. It's likely even lower today. The federal government is essentially paying the elderly not to work. Of course there would be mass hysteria and rioting if people over 65 were told they simply had to keep working.
You're not going back far enough. Elder care was not an issue that suddenly arose because of the Great Depression. It was a growing problem all over the world for decades that largely coincided with the lengthening of the average lifespan. Because lifespan was lengthening it was nearly impossible to accurately forecast financial needs into the future either for individuals or their families. In fact, the original "benefits age" for SS was placed above average life expectancy.
See:
https://aging.senate.gov/crs/aging1.pdf . Same source indicates life expectancy at birth for those born in 2003 will be approximately 77.5 years, nearly 20 years above where it was at in 1930.
Leading causes of death for retirees at the time social security was enacted included basic infectious disease that was easily curable, and infections that are largely seen in as mortal risks among those living below the poverty level (diarrhea and enteritis). Today it's largely heart disease, cancer, and stroke.
What has happened over time is that life expectancy has continued to rise and the demographics of the elderly have changed dramatically. Elderly that previously would have been dead are now alive but are too sick or decrepit to work. In 1900, only 39% of males and 43% of females ever saw 65. By 1997, those numbers increased to 77% of males and 86% of females.
https://www.cdc.gov/nchs/data/ahcd/agingtrends/01death.pdf
This is especially true in fields that require large amounts of manual labor or time spent in an environment that is not conducive to health (can you imagine a coal miner working into his 70s?).
If you look at the numbers that you provided I don't think you get the "paid not to work effect" that you're projecting out. If you dissect this out you'll see a much larger number of people at the high tail end of age (25+ years over 65) that in previous decades would have been dead and are unable to work bringing the percentage of working people over 65 down. You'll also see a much larger number of people between 65-90 that would not have made it to that age in previous generations, some of whom are working and some of whom aren't. You're also likely to see decreases in working over time simply because career shifts become notably more difficult as persons age and demographics of the working populace shift (it's easier to work as you age, for example, if you run your own small business and have control of your own means of production; also easier if the field you are working in changes less over time; both of which are less common now than ever).
There's a lot of complex demographic factors here that you're papering over by asserting a strict "paid not to work" trade-off. At best, you've provided an argument for either a) increasing the retirement age to reflect changes in life expectancy or b) changing payout levels depending on what age a person chooses to elect benefits. Both of those are probably negotiable. Becoming the only Western country in the world that displays a total indifference to elderly needs is not.
On a different note, do you realize you have unintentionally made an argument for greater government spending during periods of economic weakness by defending SS as a necessity during the great depression? Isn't that the exact opposite of your current position that we need to cut spending drastically now during a time of economic weakness?
Right now the cost of administrating SS is as high as 328,000,000,000.00 That's billions. SS paid out 439 billion in benefits. Roughly speaking, for every four dollars paid out society as a whole bears up to a 3 dollar adminstration cost.
Those are high end numbers from 2002 provided by a right-wing think tank advocating privitization (you didn't cite the source, so I had to go track it down). Those numbers are created using a set of highly biased assumptions that you can read about yourself if you go through the paper (for example, it charges significant portions of the IRS budget to the SS program as an administrative cost without describing why IRS costs decrease if SS doesn't exist).
Other estimates of administrative costs for 2009 come to $6.2 billion vs $675 billion in payouts. That's greater than 100:1 rather than 4:3.
https://money.usnews.com/money/blog...010/08/05/social-security-costs-exceed-income
CBO studies have repeatedly found that Social Security Administrative costs are significantly lower both in gross terms and as a proportion of the amount of money that comes in than alternative retirement funding arrangements. For example:
https://www.cbo.gov/ftpdocs/52xx/doc5277/Report.pdf
Page 10 of the PDF provides a nice comparison of costs. Even the CATO institute acknowledges that administration costs for SS are comparatively low because it has a built-in collection agent via the IRS and because they avoid money management fees and functions.
Imagine putting more money in every single worker's pocket as well as cutting 320 billion from the budget. At the very least SS should give workers the choice of opting out if the don't want to participate. Of course they can never do that because the government desperately needs every dollar they can get to keep the sinking ship afloat.
Actually that would just make the system less efficient, but that's a different discussion.