David R. Henderson  

Social Security Reformer

PRINT
The intellectual as a celebrit... Rio Reality Check...

The Committee for a Responsible Federal Budget has come up with a very clever tool. It's called "The Reformer: An Interactive Tool to Fix Social Security." Notice that it's to fix Social Security, not what I would like to see, which is abolition. But still, it's very clever and, assuming the underlying model is done right, it's illuminating.

So, for example, let's say that you are convinced that Social Security should be means-tested. Co-blogger Bryan Caplan advocates that and has brought me kicking and screaming to that conclusion. The Reformer gives you 3 options for doing that. I chose one that I thought people could handle: slow benefit growth for the top half of earners. Result: 35% of the 75-year gap is closed.

Another one: Index the age to longevity. Result: 19% of the gap is closed.

We're over halfway there and we haven't had to cut benefits for any current recipients who are in the bottom half.

One more reform: Index COLAs to CPI minus 1%, on the assumption that Mike Boskin is basically right that the CPI overstates inflation by about one percentage point. Result: 63% of the gap is closed.

The total is 35 + 19 + 63 = 117. In other words, the gap is more than closed so that at some point, the FICA payroll tax could be cut, possibly by a percentage point (0.5 for employer and 0.5 for employee.)

There doesn't seem to be interaction between the variables, which makes me wonder how good the model is. For example, the COLA change above should be expected to decrease the effect of indexing to longevity, from 19% to something less than 19%. I don't know how important a factor this is.


Comments and Sharing


CATEGORIES: Social Security




COMMENTS (20 to date)
Hazel Meade writes:

I think Caplan is correct that the only way out of this mess is means testing.
It can't be fixed, but it can be allowed to slowly evolve into a welfare program, which will mean acknowledging that the taxes people pay are just taxes and not investments in a retirement account.


ThaomasH writes:

But it did not include anything fundamental like funding it from consumption taxes instead of wage taxes or even uncapping the wages subject to tax. I do think raising the "retirement" age would be beneficial to help phase out this outmoded concept.

The current practice of taxing benefits makes it a sort of means tested program so I don't see anything else needed there.

I'd propose a reform of 401K type plans to give the tax subsidy as a partial credit (perhaps at higher rates fro low income taxpayers) rather than as a deduction.

Johnson85 writes:

What I don't get is why none of these calculators ever provide an option to just phase out the insane subsidies to one earner couples. Not sure how big of a dent it would make, but it's such a no brainer, you'd think it would always be one of the first options, even if it only fixed a couple of percent of the problem.

Toby writes:

David,

Regarding abolition: how do you respond to critics that point to the situation in Chile where the privatization of social security seems to have failed?

http://www.latimes.com/business/hiltzik/la-fi-hiltzik-chile-social-security-20160812-snap-story.html

That promise was that pensions would provide 70% of a worker’s final wages at work; the real figure is 38%, the lowest rate among developed countries other than Mexico.

I am in favor of privatizing pension systems in general, but from what I understood fees ate up most of the benefits. I can see this happening in the US as well due to its seeming lack of effective regulation due to rent-seeking pressures.

I am otherwise surprised that you didn't choose the option to diversify the trust fund to increase returns. That would add 21%. You then wouldn't have to index to longevity which according to some would effectively cut benefits for the bottom half who live shorter lives.

Brad D writes:

@Toby

In reality, there is no trust fund, only IOUs. For the longest time, the FICA tax brought in more revenue than expenses, leaving a surplus. Congress, unable to keep its greedy paws off the money, spent it all each and every year. This "spending" was off-budget, and an accounting trick to lower true federal outlays.

The trust fund exists only in government bonds, and social security recipients want cash not bonds. The only way to generate cash is to increase taxes or print money.

Micke writes:

@Hazel Meade:
Highly doubtful if means testing is much of a help. Means testing is very expensive even assuming an efficient bureaucracy. If it's broken, like in the US, it's probably a terrible idea.

Some actual figures from Sweden:
Child support payments (not means tested): 99% of the total cost is money actually paid out to parents. 1% is lost in overhead.

Housing subsidies (means tested, abolished a number of years ago): 50% of total cost was money paid out to beneficiaries, and 50% was last in overhead.

I seriously doubt that means testing saves anything at all on the bottom line. It will create a massive cadre of civil servants who are ever more eager to keep the system, but in my book that would be a negative, not a positive.

Swami writes:

Simple solution to the problem:

Allow people the option to choose either higher premiums or later collection date (with both options actuarially determined to be break even).

Older people would logically choose the higher premiums until they retire. Younger ones could choose whichever they prefer.

The decision doesn't have to be binary of course, their can be gradations of choices.

David R. Henderson writes:

@Toby,
I am in favor of privatizing pension systems in general, but from what I understood fees ate up most of the benefits.
I don’t know if they ate most of the benefits, but they certainly ate a lot of them. If I recall correctly, this is because the privatization law required a certain minimum expenditure on administration, including advertising.
This is one of the reasons I am against privatization. I want abolition, plain and simple.

David R. Henderson writes:

@Swami,
Allow people the option to choose either higher premiums or later collection date (with both options actuarially determined to be break even).
That might be a good idea, but if it breaks even, then, by definition, it doesn’t solve the problem.

Your idea is that we will keep the name of the program, and what it does is beside the point.

CRFB's measure is the 75 year solvency, one that understates the system's core problems because it books revenue today without recognizing the cost associated with collecting the revenue. Solvency is the definition of kicking the can. 75 years is cost to make our problem a bigger problem for our grand-children. So no CRFB's model has nothing to do with actually fixing the system.

The problem with your increase the retirement age to 69 is that you are talking about people whose retirement age was raised in 1983. That increase pretty much explains all of the increase in life expectancy during retirement until 2050. These people have paid in the most of anyone, and they have the least to do with the financial imbalances in the program. So your solution is to punish the people least responsible for the problem.

Boskins is arguing substitution. That isn't the measure of inflation. That is a measure at least in part of the behavioral response to inflation.

Here is another idea. Let's reduce the benefit of left-handed people. If that would bring your math to 100% would it be a good idea?



SS writes:

One of the ways to fix Social Security is to take the cap off the amount of money/earnings that are subject to Social Security taxes. The Social Security portion (OASDI) is 6.20% on earnings up to $118,500 for 2016. The Medicare portion (HI) is 1.45% on all earnings. Also, as of January 2013, individuals with earned income of more than $200,000 ($250,000 for married couples filing jointly) pay an additional 0.9 percent in Medicare taxes. Social Security taxes should be on all earnings too including passive earnings/investment income.

On "The Reformer: An Interactive Tool to Fix Social Security," there is no option to apply Social Security taxes on ALL earnings. Subjecting all earnings to Social Security taxes would make a huge difference.

If Social Security pensions are not considered pensions because current taxpayers guarantee/pay for the Social Security pensions, then government/union pensions should not be considered pensions either because they are guaranteed by current taxpayers.

Many government workers do not pay into Social Security because they pay into pensions guaranteed by taxpayers so why aren’t they subject to this “haircut” discussion too?

Since government pensions are guaranteed by the taxpayer, then government pensions can be reduced/means tested/subject to age restrictions, etc., like Social Security. Additionally, we could index age to longevity on government pensions. Whatever we do to Social Security pensions we should do to all government pensions. (BTW, the liability for Social Security pensions is equivalent to the liability for government pensions, but there is no discussion concerning age restrictions/reducing/means testing government pensions.)

Congressional pensions are guaranteed by taxpayers, therefore they can be reduced/means tested too and subject to whatever they decide to do to Social Security pensions.

Since the Pension Benefits Guaranty Corp (PBGC), a government organization (http://www.pbgc.gov/) uses taxpayer money to guarantee the pensions of companies who can no longer afford defined benefit plans, we could reduce/means test/increase the age required to receive the PBGC pensions/payouts, since taxpayer money is used.

We could end all taxpayer guaranteed pensions and claim it is welfare since the recipients might receive more than they put in and all of them are being supported by current taxpayers.

If taxpayers can no longer afford to guarantee Social Security pensions, then taxpayers can no longer afford to guarantee any pensions. And, since Social Security money was used by Congress for non-Social Security purposes, then our elderly/Social Security recipients should be first in line before
*The money we owe to foreign governments
*Foreign aid
*Defending any country other than the US
*Welfare, Section 8 Housing, entitlement programs for illegal immigrants
*Supplemental Social Security and Medicare/Medicaid/Disability given to elderly immigrants who have not put a dime into our system
*Etc.
Why aren’t all pensions (Social Security and government pensions) being discussed? Why are the discussions only concerning Social Security pensions taking a haircut? Why does it seem to be okay to means/age test Social Security but not all pensions guaranteed by taxpayers?

mike shupp writes:

Johnson85: To many people, single-earner households -- a working father, a stay-at-home mother and several kids -- represent an ideal form of family structure. It's the way families operated in the 1940's and 1950's and early 1960's. when Life Was Normal and Things Ran Properly. It was Virtuous. It produced Decent Kids rather than Delinquents and nobody in sight became sexual Deviants. Think of OZZIE AND HARRIET and LEAVE IT TO BEAVER, not ALL IN THE FAMILY!

Since single-earner households are so desirable, reasonable subsidies to encourage couples to form such marriages and raise children in them make sense and are a legitimate government expense.

You might want to argue with that view. Feel free! It's not great history after all, even if many elderly people remember things that way. But in a way that's my point -- family taxation policies aren't determined by purely economic considerations.

Dale Courtney writes:

Social Security is an earned benefit. I could be convinced of means testing, but only after the government pays back what it took plus interest. I think of how wealthy I would be today if my social security deductions had been placed in the stock market, or even in the bond fund.

Jim Glass writes:

I seriously doubt that means testing saves anything at all on the bottom line. It will create a massive cadre of civil servants who are ever more eager to keep the system

In this particular case means testing is easy, and it has already started, no new bureaucracy needed. Benefits are taxed on the normal income tax return based on total income reported -- with the tax on benefits returned to the SSA instead of going into general revenue.

They've been doing this since the 1983 law changes as a carefully disguised means test, making it look like just ordinary income tax. Of course, they'd have to ramp it up and make it really visible, which will be really unpopular, but there aren't many credible alternatives.

Jim Glass writes:

Social Security is an earned benefit. I could be convinced of means testing, but only after the government pays back what it took plus interest

Impossible by arithmetic, alas. SS is a paygo program, benefits paid can't exceed taxes collected overall. Thus, since early participants received much more than they paid in, today's and future participants must get back that much less than they paid in. There is no possible way around it. The SS Trustees put this deficit, the "unfunded obligation for past and current participants" at $27 trillion. That's how much more the early generations got back than they put in. (The swing year cohort that just breaks even is this those born in 1953.)

There is no possible way to close this unfunded obligation other than by cutting benefits or raising taxes paid by the younger generation, be they on payroll, income, national park admissions, whatever.

(BTW, that $27 trillion is the unfunded obligation to current participants, not looking out to the "infinite horizon" - the infinite horizon unfunded obligation actually is a little less.)

Jim Glass writes:

I am otherwise surprised that you didn't choose the option to diversify the trust fund to increase returns. That would add 21%.

That would require selling the $2.7 trillion of trust fund bonds to the public in one shot. Doesn't seem plausible. Moreover, projecting that this would produce an annual investment return of 8% annually seems very optimistic in today's reality. What would the investments be in, and at what risk? Me thinks this was a plausible idea maybe 20 years ago, too late now.

Floccina writes:

I would first cap payments above $200/week in 2016 dollars until no one gets more than $200/week. Then slowly raise everyone below that up to $200/week. Then roll the taxes into the income tax (or better yet replace income tax with a progressive consumption tax). Then it is obvious what it is, a welfare program and we can talk about whether we should keep it.

David R. Henderson writes:

@Floccina,
I would first cap payments above $200/week in 2016 dollars until no one gets more than $200/week. Then slowly raise everyone below that up to $200/week.
I was trying to work with proposals that have some chance of passage. Immediately cutting the Social Security benefits of millions of people by over $10,000 a year, which is what you propose, is a non-starter politically.

Jim Glass writes:

I was trying to work with proposals that have some chance of passage.

I think we've already seen the 'best by test' there, in what passed Congress in the test when SS went seriously insolvent and was "reformed" in 1983. Nothing indicates to me that any of the fundamental political dynamics have changed.

That bailout package consisted of three parts; (1) basic increase of payroll tax, plus (2) the start of means testing (by subjecting benefits to progressive income tax rates and returning the tax collected to SSA - a carefully disguised means test), plus (3) cutting long-term costs by reducing benefits for the young via later retirement age. At 50-50 odds I'll bet right now that these three provisions revisited will provide >75%% of the coming SS "reform" circa 2035.

The problem will be that the 2035 fix will have to be a much *bigger* hit to participants than was the 1983 reform. The tax increase then was 0.24% of GDP, the benefit cut averaged about the same going forward, call it 0.5% of GDP. This time the cash flow shortfall of SS will be 1.6% of GDP, three times larger. This is going to leave the young very visibly getting back significantly less than they put into it, which will not be popular. And nobody is going to be able to 'disguise' means testing this time.

OTOH, how is the left going to be able to resist means testing then? It's easy today to proclaim, "no means testing ever!", while the program still is at breakeven. But how "progressive" is it going to be then to argue "protect the benefits of Bill Gates and the rich by increasing taxes on workers now plus slashing benefits for the young!" I'm going to enjoy watching Krugman an his kind flip 180 degrees on this one.

Larry writes:

Rather than cut the payroll tax rate for everybody, cut/eliminate the employer side for low income workers. That adds another lower rung on the economic ladder, just what we need.

Disqus, disqus, disqus, please!

Comments for this entry have been closed
Return to top