Fixing U.S. retirement plans

There are many experts who believe that the current paradigm for retirement plans is so flawed that we should scrap it and try something radically different. Dr. Teresa Ghilarducci is one leading voice in this area, for example. The data suggesting that we need a change are compelling. Only about 50% of workers have an employer sponsored retirement plan such as a 401(k). People who do have retirement plans at work have accumulated balances that are far less than than they really need.

401(k)s, IRAs, and related plans are what are referred to as Direct Contribution (DC) plans. People put in money, sometimes bolstered by additional contributions from their employers (referred to as matching funds). The money needs to be invested and grows over time. When people retire, they can draw income from these accounts. There is no question that the current model, in which people determine whether to save, how much to save, and then must also determine how to invest, is not working. There are, however, many features of 401(k)s and IRAs that are very attractive. If we can fix some tractable problems, there is a path to radically improving retirement prospects for the American workers.

Low participation and contribution rates

The first and most significant flaw in the DC plan world is low participation rates and low contributions rates. Given the choice to save, many people simply opt out. Many people save so little that a large amount of the matching money offered by employers goes unclaimed. In 2015, for example, one analysis covering 3.5 million workers concluded that employees were losing a total of $24 Billion, or an average of $1,336 per year, because they were not contributing enough to take advantage of the full match.

There is already a tried-and-true solution to the issue of low participation and contribution. Australia has what is called the superannuation plan, a DC model, for which participation and a specific level of contribution is mandatory. Employers are required to contribute 9.5% of an employee’s base salary and employees can make additional tax-advantaged contributions if they wish to do so. Moving to this kind of model goes a long way towards fixing the problem of participation and low savings.

What about for people who are self-employed or who, for some reason, don’t have access to a plan through their employer? The best solution is to facilitate the creation of IRAs or related self-directed retirement savings plans that are tied to the individual rather than the workplace.

High costs

The second problem with DC plans is high costs. Some DC plans have all-in costs well above 2% of assets per year, which is very expensive. The largest plans, typically at companies with large numbers of employees, have the ability to negotiate more effectively because of their larger size as well as leveraging economies of scale. Quite a few have plans with total costs of around 0.5%.

Total costs of 401(k) plans by size (Source: Brightscope)

Small plans will always be somewhat more expensive on a per-worker basis because of economics of scale, but we need to figure out how to make 401(k) fees lower for small and mid-sized firms.

The potential to lower costs is demonstrated by Solo(k) plans, which provide self-employed people with all of the benefits of 401(k)s. Solo(k)s are incredibly inexpensive. Granted, companies have additional costs to maintaining plans, but there should be room for substantial improvement. There is no reason why the Solo(k) model could not be applied to companies, such that a company would contribute to a Solo(k) for each employee and employees could contribute additional money if they wish.

Another way to lower costs would be to expand the Thrift Savings Plan (TSP), which is functionally identical to a 401(k) plan and is offered to federal employees. The all-in cost of TSP is 0.4% per year. Just like a 401(k), TSP provides access to a range of investment funds. There would be potential problems with a massive expansion of TSP, even if this was cost neutral to the taxpayer. Specifically, this kind of program would potentially drive massive well-established firms (that currently provide 401(k)s) out of business or into dire straits, resulting is job losses, etc. This would amount to nothing less than nationalizing 401(k) plans.


If we moved to a system of DC plans like 401(k)s, but with required employer contributions and with substantially lower costs, Americans would be far better off. If we could reduce costs with an offering similar to the Solo(k) plans available to the self-employed, this would be a massive improvement in for many millions of Americans. This account would simply follow the employee from job to job and seamlessly supports self employment. There is nothing new to invent or prove. Australia has shown that requiring employer contributions works well. Further, the fact that so many financial firms offer low-cost Solo(k) plans and IRAs shows that these products are financially attractive to these firms, even with their low costs. As Americans move through a series of jobs and more people are self-employed, a Solo(k) for all workers (with required employer contributions) seems like a very good solution.

Given that both Social Security and many public pensions have far less in assets than they need to meet their commitments, many people will probably be happier with owning the underlying assets, rather than having just a promise of future benefits. While Social Security is a very important program for many Americans, the simple reality is that younger people cannot trust that they will receive what they are currently being told that they are estimated to get. As the Social Security Administration writes on everyone’s annual statement (

Your estimated benefits are based on current law. Congress has made changes to the law in the past and can do so at any time. The law governing benefit amounts may change because, by 2034, the payroll taxes collected will be enough to pay only about 79 percent of scheduled benefits.

The good thing about DC plans is that your assets are, in fact, yours. Relying on a promise of future benefits, such that these benefits may or may not actually get paid, is looking fairly shaky at this point. I am not counting on getting all of the benefit that Social Security is telling me to expect, although I do think that I can probably expect around 3/4 of this amount. That said, the text from SSA does not make me feel highly confident.

In summation, then, DC plans can be fixed using a combination of required contributions and making these plans into low-cost, low-overhead offerings more like Solo(k)s. Given the state of traditional pensions, whether public or private, I believe that Americans will be better off owning their retirement assets rather than relying or a promise of future benefits.

One thought on “Fixing U.S. retirement plans

Comments are closed.