Ezra Klein's Fudging of Social Security Facts and Figures

Ezra Klein is one of my favorite policy wonks. I read him on a regular basis, and will continue to do so. But today, I must set him straight. In a piece criticizing Alan Simpson - who, yes, I agree is guilty of some pretty bad insensitive language - Ezra Klein takes refuge of an interesting statistic: that for people who reach are Medicare age, life expectency has only increased by 5 years since 1935.
In 1935, a white male at age 60 could expect to live to 75. Today, a white male at age 60 can expect to live to 80.
Notwithstanding the fact that the workforce of this country has diversified quite a bit since 1935, both in terms of gender and race, this has been the line for people trying to counter the argument that since social security's inception, general life expectancy (for people who are born in a given year) has increased dramatically. To be sure, for someone born in 1935, the average life expectancy in the US was about 60 years - below the social security retirement age of 65. In 2006, it was 77.7 years.

In terms of social security solvency (when we say solvency, we are generally talking about a 75 year solvency of the social security trust fund), we are not simply talking about people who, as Klein seems to imply, reach age 65. As with the solvency of any fund, we are talking about the overall intake and output of the program. It's not exactly an easy argument to make, but the dispassionate fact is that in the early days of social security, even those who passed away before they reached the age to collect social security benefits paid into to social security, thus leaving a relatively large fund to draw from for people who did reach retirement age. The fact that a lot more people died (proportionately) before they could collect social security helped tip the balance in favor of a social security surplus. Again, it's not a convenient argument to make, but it's the truth.

But life expectancy at birth is not the exact right measure to determine the proportion of people who got to collect social security vs. those who did not. Because children in general do not generate social security income or taxes. So let's look at working age Americans - life expectancy at age 20, i.e. once someone turns 20, how much longer they are likely to live.
  • In 1939, the average 20 year old was likely to live another 48.5 years, meaning they would collect social security, on average, for 3.5 years.
  • In 2006, the average 20 year old was likely to live another 58.6 years, meaning they will collect social security, on average, for 11.6 years (adjusting for retirement age increase to 67 years).
So the average retiree may only collect 5 years more of benefits compared to 1935, but that is a deceptive figure, since a far greater percentage of the population today are likely to reach retirement age than were in the mid to late 1930s. The truth is that today's 20 year old, on average, will collect social security benefits for 8 years longer than a 20 year old in the 1930s.

That by itself may not seem much (5 vs. 8 years) but that's not the full story. Add to it the fact that Americans are having less children, thereby skewing the population numbers to the elderly. Proportionally, today, we have 3 workers for every social security beneficiary. In 1950, the ratio was 16.5 to 1. When we are talking about numbers, we need to examine their full context.

Here are a couple of other facts and figures that Ezra Klein uses that serve to mislead. Klein tells us how the yearly earning replaced by social security is on the decline:
For the average 65-year-old retiring in 2010, Social Security replaced about 40 percent of working-age earnings. That “replacement rate” is scheduled to fall to 31 percent in the coming decades.
See, this is yearly earnings. And indeed, the average replacement rate is falling since incomes are projected to increase faster than social security benefits, and in the progressive benefit formula, the less you earn, the more of your earnings are replaced by social security (as it should be). This is a fine statistic to present, I guess, but it's important to remember that here, Ezra Klein is talking about yearly earnings/benefits.

In this light, it's rather fascinating that Ezra would then turn around and present this number:
Raising the retirement age by one year amounts to roughly a 6.66 percent cut in benefits.
In lifetime benefits. But you cannot say this without recognizing the flip side of this same coin: for every year lived longer, one's lifetime social security benefit increases by 6.66%. Which means that according to his calculations - a 5 year increase in longevity for people who reach age 60 - the average lifetime social security benefit is now now 33% higher. By my calculations, every 20 year old entering the workforce today can expect to have an increase of 53% in their (lifetime) benefits than those who entered the workforce in 1939.

Suddenly, Ezra flips around and starts using overall lifetime benefits calculations when he's talking about raising the retirement age, but talks about yearly income when talking about the "replacement rate." I'm sorry but you cannot have it both ways. Either calculate the "replacement rate" as a ratio of total lifetime benefits vs total lifetime earnings (within the Social Security cap), or admit that raising the retirement age has no effect on yearly benefits. To do otherwise is not intellectually honest.

The trust fund's funding problem is compounded by three things: people living much longer, the ratio of workers to beneficiaries falling, and the cap progressively taxing a smaller portion of the national income as the gap between the super rich and everyone else widens.

So what do we do? Dramatically raise the retirement age? That doesn't seem plausible at all. In his piece, Ezra also ignores that the Fiscal Commission Chairs proposal (Simpson was one of the co-chairs) includes many of the solutions to the long term solvency of social security that Ezra Klein himself proposes. Let's look at those:
Social Security’s 75-year shortfall is manageable. In fact, it’d be almost completely erased by applying the payroll tax to income over $106,000.
Quite true. This would be my preference as well. However, originally, social security was structured to tax 90% of the national income. Today, thanks to widening income inequality, that percentage has fallen to about 82%, as I wrote last year in my series about the Fiscal Commission Chairs' recommendations. Alan Simpson and Erskin Bowles (the co-chairs) proposal does include a provision to restoring this rate back to 90% of national income, and ensures that it stays at that level.

More from Ezra's solutions:
Most opinion elites — Simpson being one good example, and the U.S. Senate being another — show a very strong preference for working as long as possible. Most Americans show a very strong preference for retiring as early as possible. Elites who enjoy their jobs need to be very careful about generalizing their experience to people who don’t enjoy their jobs. More bluntly: Raising the retirement age is the worst of all possible options for reforming Social Security. It’s not only regressive, but it also falls most heavily on those with the worst jobs. Means-testing would be much better.
I hope that Ezra Klein isn't seriously arguing that people ought to be able to retire, at whatever age, simply because they do not like their job, and be eligible for Social Security. Ezra, there is also means testing in the Bowles-Simpson proposal.
It establishes a new special minimum benefit, defined above the poverty level, for the workers who worked the hardest for the least amount of reward. This is a very welcome step. The proposal also wage-indexes the special minimum benefit, ensuring it stays relevant in the future, and the floor on benefits has a basis in wage. It also suggests a benefit boost to older retirees most at risk of outliving other retirement resources.
There is also means testing in terms of the toughest jobs. While the proposal raises the retirement age to 68 in 2050 and 69 in 2075 (tied to longevity trends, not absolute), there is a hardship exemption for those who are unable to work beyond age 62. At age 62, these workers would be able to get the full social security benefits, not reduced benefits as one would get by retiring at age 62 today.

I understand that Alan Simpson has a foul mouth. I understand he's said some pretty crazy things. But Alan Simpson is not - or at least should not be - the topic of discussion here. The topic is social security. It's about the plan the Fiscal Commission co-chairs presented for social security. Everyone should be united in the belief that social security must remain solvent without burdening those who can least afford it, and the wealthy must pay a fairer share (and frankly, I don't mind much, and neither, it seems, does Ezra if the benefits for the rich are reduced). What we ought not be doing is using statistics in ways that fool people - by using lifetime benefit data for one thing and yearly data for another. Let's have a sensible discussion, please.

I respect Ezra Klein. I like what he has to say. But I must hold him accountable as well. As progressives, we should not stand for the Right wing scaring everyone about the viability of social security - we know that the long term funding issues can be addressed with good, progressive reforms. But as progressives, we also should not be trying to hold up a boogyman (Alan Simpson) and try to gin up visceral reactions against him instead of focusing on the issue.