Pension/retirement arrangements where the contributions are defined but the benefits are not are a risk to the individual -- you have to figure out "how much do I want to invest, what do I think the market will do, where should I put the money?", and if you get it wrong you don't have enough money in retirement. For the company, on the other hand, they're very low risk, because the amount of the company's liability is clearly defined and limited up front.

Arrangements where the benefits are defined are not a risk to the individual (who knows how much they will get, and will get it regardless of market conditions over the next few decades); instead they're a risk to the company/government/etc that provides them (if they set the defined final pension too high and investments don't go so well then it costs the company more than they expected).

In general my preference is that (as a society) we should prefer to impose risks not on individuals but on large organizations who can afford to employ people who take the time to assess those risks and make the necessary decisions, and who have the resources and timescales to be able to weather unexpected downturns. So I think defined-benefit is better than defined-contribution.

(This is of course entirely not the way the pension setup has gone, at least in the UK -- defined-benefit is getting rarer and rarer, and defined-contribution is pretty much the standard for the private sector these days. But I think it's an unfortunate trend.)

Like the whole company? Only a government that can print money can really offer such a defined benefit, and even then it is just a mirage since the actual value can be inflated away. The real issue is that it is actually impossible to make such guarantees since conditions can always change, but some people don't want to accept that reality.

I think this is slightly overstated. The most common outcome in this situation is that you work more, the amount of extra work being determined by the size of the error.

> In general my preference is that (as a society) we should prefer to impose risks not on individuals but on large organizations who can afford to employ people who take the time to assess those risks and make the necessary decisions, and who have the resources and timescales to be able to weather unexpected downturns. So I think defined-benefit is better than defined-contribution.

This isn’t a bad point, and is strengthened by the observation that pensions benefit from pooled risk. E.g. those that live longer are subsidized by those who unluckily die younger, so everyone on average gets away with slightly lesser contributions (or earlier retirements).

But it misses out on incentive mismatches with pensions. A pension administrator is probably not going to work at the company in question for their whole career, and their incentive career wise is probably not the long term maximization of risk adjusted return for this pension, but instead perhaps taking risky bets hoping they pay off for notoriety, or avoiding risk at all costs to avoid complaints.

And this dovetails into some really hairy aspects of pensions that aren’t often brought up.

For one, they generally punish people for alternative lifestyles. Imagine trying to be a childless person retiring early with a pension that doesn’t kick in until 65 because the typical person has kids and works that long.

Also, pensions can be discriminatory. Imagine having a disease that shortens your lifespan, like MS, or a higher risk factor for something like early onset dementia. It’s entirely possible you’re forced to contribute to a pension program you have every likelihood of not being able to benefit from in your natural lifespan.

For a more common example, women live longer than men. Should men have to contribute less for the same benefits then?

Pensions work just the same as 401ks, but better when properly funded and governed, because your average worker is unsophisticated as it relates to investing. 401ks were a scam foisted on the public as pension replacements, with corporations instead juicing shareholder returns with what was previously pension contributions. If you're a high earner without any life events that would cause you to liquidate your 401k (extended job loss, medical event), congrats! You have won the lottery.

https://www.nbcnews.com/business/retirement/great-401-k-expe...

https://www.economicpolicyresearch.org/images/docs/research/...

https://www.gao.gov/financial-security-older-americans

https://web.archive.org/web/20201026175751/https://www.gao.g...

https://www.empower.com/the-currency/life/average-401k-balan... (direct your attention to median balances)

https://www.ojp.gov/ncjrs/virtual-library/abstracts/invisibl...

https://www.dea.gov/sites/default/files/pubs/states/newsrel/...

I would much rather have my retirement funds in a 401(k). The assets are in my name and can't be arbitrarily taken away.

That wording implies voluntary.

The new loyalty relationship is immensely asymmetric: "the company" holds you by the heart plug (health plan, visa, vesting) while at the same time will boot out a substantial fraction of its workforce, regardless of merit, the moment a quarter turns yellow in some spreadsheet or the M+A pendulum comes around again.

There are very few places anyone can spend their whole career at one company even if they wanted to. I've been at this 30 years and I think I've had 10 jobs but only quit once.