Friday, August 28, 2015

Smith: “Social Security is not an anti-poverty program”

Brenton Smith, writing for The Hill, states that

“One of the most pervasive myths in the debate about Social Security promotes the role of the program in the alleviation of poverty. Common sense should tell us that something is amiss with this endearing myth. Social Security does not pay a penny of benefit based on need.  The system does not even have visibility into need. So at best any benefit that goes to a poor person is more a matter of luck than systemic policy.”

I don’t agree with all of his argument, but he makes a point: as a social insurance program, Social Security is far from perfect. Check out my 2009 paper, “Will Your Social Insurance Pay Off? Making Social Security Progressivity Work for Low-Income Retirees.”


JoeTheEconomist said...


I welcome your thoughts on the argument. I freely admit that I have no idea how the system interacts with poverty. Word count stopped me from talking about capital appreciation, which isn't counted. So income is even more understated than I mention.

I know that you favor making SS more about poverty alleviation. The problem with progressivity is that it deals with past earnings not current wealth.

The problem is that the system is completely unsuited for the job. End the system, and transfer the assets to a program that actually does work with poverty. When you only have a hammer in the tool box everything looks like a nail.

Andrew G. Biggs said...

To directly target poverty you'd need to use a means-test, which has some significant downsides:

You can avoid some of them by targeting benefits to something that's correlated with poverty in retirement. My guess is that the closest correlate would be lifetime earnings, so arguing that Social Security doesn't target poverty at all seems wrong to me.

That said, I don't think the program does as good a job as it should, which is why I've argued for reforms that attack the issue a bit more directly:

Arne said...

"lost roughly $600,000 in savings"
should be
lost roughly $600,000 in POTENTIAL savings

Understanding how SS interacts with poverty would require understanding people's actual saving behaviors.

"more a matter of luck than systemic policy"

Depending on an obvious correlation is hardly luck.

"Poverty should be measured by net-worth"

I would say poverty should be measured by spending potential. A 62 year old with $600K, about $38K per year. An 80 year old, much more. Over 65's may be the wealthiest group, but that wealth drops rapidly with age. I am pretty confident that there is a strong correlation between wealth and income.

My critique aside, I do think that folks should be thinking about TANF and Supplemental Security Income more than Social Security when they think anti-poverty.

WilliamLarsen said...

taxable income in retirement is not a good indication of poverty.

The entire purpose of saving for retirement is to create a nest egg large enough to fund your retirement. That being said,that nest egg is sold off periodically to fund an individuals life style. In the first year of retirement, your nest egg creates the largest amount of income (unless the market is in a down turn). This would mean most of your needs come from income and a little from principal.

Over a 22 year retirement period how much of your needs is provided by principal and how much by income? In the first year 98% of your needs come from income and 2% from income. In year about 67% is from income and 33% from principal.

By year 20, 25% comes from income and 75% from principal.

Measuring income alone does not provide much indication of poverty since half of all needs over 22 years comes from principal. What I am attempting to say is that if a person in year 20 says they have $10K in income, they should be getting $30K from principal for a total of $40k. Is this person living in poverty because their income is $10K?

The only sure way is to look at assets. That then becomes a problem. If there assets are in a home, it generally does not provide income and would show they live in poverty while living on a 100 acre prime farmland or a $1 million home. Is it wise to all non liquid assets not to be counted? Non liquid assets are considered by some financial planners as "quality of life" assets. Do we want other taxpayers living in with a lower standard of living to pay benefits to a person who is living better?

SS is facing the problem that FDR wanted to avoid at all costs.

JoeTheEconomist said...


The people at CATO have the same problem with the 600,000 in lost savings. They say it assumes overly conservative investment options. So the potential lost savings is much larger. I am trying to use an apple to apple comparison of forced insurance to forced savings.

Statistically the spousal benefit which is a major benefit falls with a bias to the wealthy. Wages and job history fall to the wealthy. While I have seen contradictory evidence, the longevity bias falls to the wealthy. Which correlation are you talking about?

Again, I don't know the connection between Social Security and poverty. The way we measure it today is not sound.