Wednesday, December 31, 2014

The CBO weighs in on the “replacement rates” debate

On December 18 the Congressional Budget Office released its latest report on Social Security. In addition to updating its projections for system financing – which has worsened considerably in recent years – CBO touched on the debate over how to measure Social Security “replacement rates.”

Back in July, Syl Schieber and I wrote in the Wall Street Journal that the Social Security Administration’s method of calculating replacement rates understates the adequacy of Social Security benefits. A replacement rate is designed to measure the degree to which retirement income can “replace” working-age earnings and thus allow retirees to maintain their pre-retirement standard of living. Most financial advisors recommend a replacement rate from total retirement income of about 70%, while SSA’s actuaries calculate a typical replacement rate from Social Security benefits of about 40%.

But SSA’s actuaries use a method that overstates individuals’ pre-retirement earning and thus understates their replacement rates from Social Security. SSA compares Social Security benefits to the wage-indexed average of the retiree’s highest 35 years of earnings. This “wage indexing” adjusts past earnings for substantially more than the rate of inflation and thus understates the degree to which Social Security benefits help retirees maintain their pre-retirement standard of living. Here’s how we put it back in July:

“Say you are retiring at age 65 this year and earned $20,000 in 1985. The purchasing power of that 1985 salary in 2014 dollars is $43,640. But in calculating replacement rates, SSA wage indexes that $20,000 for the growth of the economy, and so under that model you earned $53,281. Replacing 70% of $53,281 is a lot more difficult than replacing 70% of $43,640. SSA’s wage indexing of past earnings in effect credits retirees with salaries that they never had, then deems retirement income inadequate if it fails to replace that nonexistent past salary.”

In its new report, CBO seems to grasp these points. CBO calculates replacement rates relative both to wage-indexed earnings (its previous practice) and inflation-indexed earnings. In both cases, CBO calculates replacement rates using the 35 highest years of pre-retirement earnings.

Moreover, the language CBO uses makes clear they understand what we were getting at:

“Indexing earnings to prices better captures the real amount of resources available to a worker over his or her lifetime, whereas indexing earnings to wages may overstate those amounts.”

How much does this matter? A lot. CBO’s wage-indexed replacement rate for a typical worker born in the 1980s is 46%, while that person’s price-indexed replacement rate is 61%, about one-third higher.

A replacement rate is merely a shorthand for a much more complex “life cycle” calculation in which individuals try to maximize their standard of living over their full lifetime, while accounting for changing family sizes, uncertainty regarding their life expectancies and the return they can receive on their savings, their attitudes toward risk, and other factors. No shorthand will be perfect.

But CBO’s use of the inflation-adjusted average of the highest 35 years of earnings isn’t a bad shorthand. It will tend to reflects earnings from the individual’s late 20s through retirement, a period in which he or she has paid off some debts, established a career path and begun to anticipate the standard of living they will enjoy through their working life, which is the standard of living they will seek to replicate through their retirement savings. If you want a single measure that’s easy to understand while being consistent with the more sophisticated life cycle approach to retirement saving, CBO hasn’t chosen a bad one.

By contrast, SSA’s “wage-indexed” measure isn’t consistent with a life cycle approach. In the life cycle model, for which Franco Modigliani won the Nobel Prize, individuals care about smoothing their own standard of living from year to year. SSA’s actuaries, by contrast, assume that individuals want their standard of living to rise each year with the average wages of other workers in the economy. But they don’t present any research or data to back this “Keeping up with the Joneses” theory, they merely assert it. Personally, I’ll stick with the guy who’s got the Nobel Prize…

1 comment:

WilliamLarsen said...

What is the purpose of Social Security? What is the criteria for determining the benefit level? How much should Social Security Pay? What should the tax rate be for Social Security? How old should one be before they can collect social security? How should social security benefits be calculated? Should social security be based on a workers contributions or should they be based on a workers wages?

There are many questions and when wage indexing is brought up, no one ever asks any of he above questions. Age, wages, tax rate base, retirement age all determine OASI revenue. However, what was the criteria used for determining the OASI benefit?

Initially they bent over and pulled a number out of someone's behind. There was no mathematical determination as to what the benefit should be. Politicians are great at doing this.

The first benefits were based off a congressional vote each year to change the table for determining the worker's benefit. When OASI was clearly not solvent in the 50's, changes were made. However, these changes still did not solve the solvency problem. The benefit became a hot potato with congress. To remove congress from the mess the 1977 OASI benefit formula was created.

The 1977 benefit formula is a good start at being fair in terms of a formula that balances life time wages with a life time OASI benefit. However, even the 1977 formula does not adjust for increase in cohort life expectancy or the increase in the full retirement age (1983) to 67.

I do not believe the 1977 benefit formula was to solve the solvency problem, but simply to create a more fair of determining an initial benefit. This formula can be actuarialy correct if the tax rate, retirement age, cohort life expectancy and inflation were all designed to be factors.

However, as will all government programs, little thought as to how all the variables would relate were never factored in.

Now wage indexing is portrayed as being "isn’t consistent with a life cycle approach" or ""method of calculating replacement rates understates the adequacy of Social Security benefits" or "SSA’s actuaries use a method that overstates individuals’ pre-retirement earning and thus understates their replacement rates from Social Security."

Before one can determine if wage indexing overstates, understates life style, one must first answer what is Social Security supposed to do? This question has never been defined in terms of replacement rate, retirement age, inflation, standard of living, cohort life span to name a few variables. Simply put Social Security was not designed to work; there was no design and that is why after 78 years Social Security simply continues to get worse and worse for each succeeding cohort.