The Connecticut legislature is considering a bill that create a publicly administered retirement plan that would be open to anyone who works at a company with more than five employees. Employees would, by default, be enrolled in the plan (at a contribution rate to be determined), but could choose to opt out. The money would be pooled in a trust, but each participant would have an individual account in that trust, and the rate of return on that account would be specified each December for the following year. Upon retirement, the account balance would by default be converted into an inflation-indexed annuity, although participants could request a lump-sum deferral.
The legislative session ends in less than two weeks, and while the bill has passed through committees, I believe it's not certain whether it will be put to a floor vote. On Friday I wrote on op-ed for The Connecticut Mirror about the bill.
The requirements for businesses are relatively modest. The main one is that they have to allow employees to make contributions by payroll deduction. If they don't outsource their payroll (which most companies do), that will involve a little more administrative hassle, but nothing terribly complicated. (Employers already have to deal with federal and state income tax withholding, payroll taxes, workers' compensation insurance, and unemployment insurance contributions, to name a few.) And compared to a 401(k), this plan has the huge benefit that the employer isn't a fiduciary. Retirement benefits are paid solely out of participants' contributions and investment returns, so there's no fiscal impact.
I think one controversial aspect of the bill, at least for some people, is the annually pre-specified rate of return. This is nothing new, by the way. TIAA-CREF, for example, has an account in its retirement plans that pays a pre-set rate of interest on a quarterly basis. But the downside is that if the interest rate is going to be guaranteed, it can't be that high. In this respect the accounts will be similar to the myRA, which will pay an interest rate based on Treasury yields.
But if you want a guaranteed return, this is the way it has to be. One of the central rules of finance is that if you want a higher expected return, you have to be willing to take on more risk. So the underlying question is how much risk people without a lot of income should be taking in their retirement accounts. (For the most part, if you have a lot of income, you already have an employer-sponsored retirement account or a SEP-IRA if you're self-employed.)
This is a question on which smart people differ. The argument for stocks is that they have a higher expected return, but that means they also have higher potential losses. While stocks tend to do better than other asset classes over thirty-year periods, there's no fundamental reason why they have to (and there have been twenty-year periods during which they haven't been the top asset class). My general thinking is adopted from Zvi Bodie: In finance, you should match your assets to your liabilities, and to the extent that your liabilities are fixed (i.e., the minimum you need to live in retirement), your assets should be as safe as possible. In addition, while stocks can be a better investment, they introduce countless possibilities for making bad investment choices, and human beings have tried out all of them.
In any case, I think SB 249 is a modest step forward. The underlying problem, unfortunately, is that many people just don't make enough to save enough money for retirement. But that's a much harder problem to solve.