Wednesday, December 16, 2015

New article: “Investment Returns: Defined Benefit vs. Defined Contribution Plans”

The Center for Retirement Research at Boston College has released a new Issue in Brief:
“Investment Returns: Defined Benefit vs. Defined Contribution Plans”
by Alicia H. Munnell, Jean-Pierre Aubry, and Caroline V. Crawford
The brief’s key findings are:

  • The analysis compares returns by plan type from 1990-2012 using data from the U.S. Department of Labor’s Form 5500.
  • During this period, defined benefit plans outperformed 401(k)s by an average of
    0.7 percent per year, even after controlling for plan size and asset allocation.
  • In addition, much of the money accumulated in 401(k)s is eventually rolled over into IRAs, which earn even lower returns.
  • One reason for the lower returns in 401(k)s and IRAs is higher fees, which should be a major concern as they can sharply reduce a saver’s nest egg over time.
This brief is available here.


WilliamLarsen said...

Sylvester J. Schieber makes a good argument that the cost of fixing Social Security should be born by all generations. The problem is that "Fixing" is a lot different from "Solving."

I do not believe social security can be fixed. This is just re arranging the deck chairs on the SS Titanic. The underlining problem is that future workers will forever pay more for their SS-OASI benefits than they are worth to any cohort as a group.

Solving SS-OASI's problem would be to basically start over. Like a junk worn out car, it is time to trade it in. Learning from the mistakes of the past;
No design criteria for how to balance benefits with taxes paid as a cohort.
Politicians setting the tax rate without regard to benefits, life span or gender.

The article states that "the lifetime value of his benefits at retirement would be 78% of the lifetime value of the taxes he’d paid on his earnings." This is based on full scheduled benefits being paid. What happens when the payable benefit is 75%? Does 78% become 58.5%? Of course the 75% is shortly after the trust fund is exhausted and as the worker to beneficiary ratio continues to drop so does the payable benefit.

The only way to increase the payable benefit is to increase the number of workers. This means Social Security is based on something that cannot be controlled. Therefore, why not base it of the actual workers contribution?

Young people do not vote because they do not understand the situation. However, with every generation their is a catalyst. Congress is going to have to do something and when it does there will be a lot of debate. Will politicians be able to keep young people from hearing the bad news that they will have to shoulder the cost?

In my local paper, there was an article about SS in response to an older individual - no I did not write it. It is drawing some attention. A few tweets, face book comments and who knows, maybe the millennials will vote with their older generation who combined out number seniors and boomers together. The article alluded to this.

Bruce Webb said...

Odd. Rent seeking by investment advisers reduces ROI from 401(k)s turned IRA's more than defined benefit plans run by legal fiduciaries.

Who would have known that the more fully informed actors in a market system would have sought to maximize their own interests absent legal compulsion otherwise?

Happy Holidays Dr. Biggs! It has been a long time.

Bruce Webb said...

"The only way to increase the payable benefit is to increase the number of workers. This means Social Security is based on something that cannot be controlled"

This can be controlled by immigration policy. Which doesn't start with a "yuge wall" even one with a golden gate that admits only high tech workers with H1B visas.

Want more workers? Relight Lady Liberty's torch. The notion that America's fundamental problem is an aging population and a declining worker/retiree ratio only make sense within a nativist context.

This is doubly true for those who wail about 'fertility ratios' while bemoaning 'anchor babies'. There is of course a way to thread this particular needle. But it doesn't read very well in languages other than the original German.

Andrew G. Biggs said...

Bruce, the CRR paper doesn't seem to show how management fees have changed over time, but fees for 401(k)s have dropped pretty substantially due to increased use of life cycle/index funds, while fees for DB plans have risen as they've shifted into alternative investments. I'd guess that today they're very close to each other.

Immigration is a tricky issue for Social Security because earnings of immigrants are substantially lower than the AWI, so it's not as simple as thinking about worker-beneficiary ratios. New immigrants provide a cash influx up front, because they're paying taxes but not yet receiving benefits. Over their lifetimes, however, most legal immigrants receive more in benefits than they pay in taxes. The overall effect on Social Security financing depends upon the earnings of immigrants, their fertility, the earnings of their kids, the number of immigrants in following generations, etc. OACT says the net effect is positive. It may be, but I suspect it's closer to a wash than they conclude.

JoeTheEconomist said...

Asset management fees are only a portion of the overall fees associated with 401Ks. Yes, the mutual fund fees are falling, but the larger problem is the hidden fees which vary depending upon who you work for.

I wrote an article on the problem about 2 years ago. Employers negotiate the terms of the 401K on behalf of their workers, which presents a massive conflict of interest. The 401K administrator charges an 'administration fee',that means the .25% can inflate by 8 fold. I think that the average fund charges 1.25% in overall fees.

Andrew G. Biggs said...

I agree on the conflicts of interest. But according to ICI (who I trust), the median "all in" 401(k) fee in 2013 was about 0.67% of assets managed.