Writing for MarketWatch, Alicia Munnell of Boston College explains why the long-term Social Security deficit has been increasing, arguing that it’s not that Social Security’s financing has been worsening so much as that the way we’re measuring Social Security’s finances will naturally lead to a larger deficit over time.
“In 1983 — the last year for any major legislation — the Trustees projected was a small surplus over the 75-year period (1983-2057). Almost immediately after that legislation, however, deficits appeared and increased markedly in the early 1990s, then dipped for a while, and then rose to around 2.7% where it has remained in the last six years.”
“The question is why the program moved from a 75-year surplus of 0.02% of taxable payroll in 1983 to today’s 75-year deficit of 2.83%. As shown in the table below, the major reason for this swing is the impact of changing the valuation period. That is, the 1983 report looked at the system’s finances over the period 1983-2057; the projection period for the 2017 report is 2017-2091. Since Social Security costs are rising with the retirement of the baby boomers, each time the valuation period moves out one year it picks up a year with a large negative balance. This moving the period forward is responsible for the bulk of today’s deficit — 1.97 of the 2.83% of taxable payroll.”
Munnell is correct, but I think she misses an important point. Policy analysts have long been aware of this “measurement window” problem: if you measure Social Security’s finances over the next 75 years, then with every passing year you’ll be picking up a new 75th year. And since those distant years are ones with big deficits, each year the 75-year Social Security deficit will increase. So in that sense, a rising Social Security deficit is entirely predictable.
The problem is that we don’t have to measure Social Security’s finances over only 75 years. Beginning in 2003, Social Security’s Trustees included in their report a measurement of the program’s finances over the “infinite horizon.” This found a substantially larger long-term long-term deficit – 3.5% of payroll in the 2004 Trustees Report, versus 1.8% over 75-years.
But the infinite horizon actuarial balance has the advantage of not rising simply due to the passage of time. So the current infinite horizon shortfall of 4.2% measured in the 2017 Trustees Report represents changes not to the measurement period, meaning either negative experiences since 2004 or more pessimistic assumptions about future economic or demographic variables.
So why don’t policymakers rely principally on the infinite horizon measure? It’s mainly not because it’s less accurate: most of the supposedly “infinite" funding shortfall is actually accounted for by things that already have happened or are predicted to occur during the next 75 years. The main reason policymakers don’t use the infinite horizon figure is that it makes the Social Security funding problem seem larger, and a larger problem is harder to solve. And elected officials don’t like hard problems.
Posts
7 comments:
Idiots is all I can write. 75 year time horizon includes the following:
All SS revenues from taxation of wages and benefits between 2017 and 2092
All SS Benefits paid between 2017 and 2092
The value of both trust funds in 2016.
What it does not include:
Those benefits accrued (earned) by workers between 2017 and 2092, but paid in year 2093 and beyond.
The benefits of anyone born in 2025 and after.
How do you keep the 75 year time horizon and correct it for those costs not included?
Eliminate the revenue from those born in 2025 and after from the calculation.
Eliminate the revenue from those who reach full retirement age prior to 2092, yet who have not reached their 82 Birthday by 2092.
What does this do?
First it eliminates the largest revenue source expense from the calculation that fall outside the 75 year valuation point. This is not a infinite time horizon calculation where you must assume hundreds of years into the future.
Eliminating 67 cohorts who contribute to SS, but receive no benefit is the correct thing to do.
For those who reach full retirement age prior to the end of the 75 year solvency period (2092), there must be some cut of date. The average life expectancy at full retirement age is about 21 years. So a more accurate analysis would add costs that fall outside the 75 years for about 21 cohorts.
Of course there is a more accurate method that could be used. With today's faster computers, an actuarial analysis of all SS wages of those still a live could be done for each person (very simple and straightforward calculation) and sum the actuarial value of each individual's wages and those taxes yet to collect.
One might think this is a time consuming task, but think about how much information the IRS processes in a given year? The sole purpose of the SSA was to record individual wages. Even if there were 300 million accounts still active that have wages being reported on or benefits being paid (there are not) it is a simple calculation.
Computers are great at repetitious calculations.
The sole reason this is not done is to keep the ponzi scheme going as long as possible. Can you imagine the anger the populous would be if the cost went from $12.5 Trillion to $35 Trillion plus? The magnitude of the tax increase is huge. The size of the shortfall in 2034 (most likely not the right year) is projected at 25%. This means tax revenues from payroll taxes will cover 75% of scheduled benefits. However, this means a tax increase of 33% just to cover 100% of scheduled benefits in 2034. Then continued increases there after until 2060 when the sum total of the tax increases will be close to 75%.
Politicians want to get elected and re-elected and telling the truth about SS is not going to win you any voters.
The main reason policymakers don’t use the infinite horizon figure is that it makes the Social Security funding problem seem larger, and a larger problem is harder to solve. And elected officials don’t like hard problems.
It is not it seems larger, IT IS LARGER! Do not placate the problem by using words like "seem."
Some annual growth rates.
The formula is Fv = Pv x (1 + I)^n
Fv / Pv = (1 + I)^n
ln(Fv / Pv) = ln(1 + I)^n
ln(Fv / Pv) = n x ln(1 + I)
ln(Fv / Pv) / n = ln(1 + I)
exp[ln(Fv / Pv) / n] =(1 + I)
exp[ln(Fv / Pv) / n] - 1 = annual rate of growth
Annualized
Year Growth (2000 - 2016)
GDP (millions) 3.78%
National Debt 8.05%
SS Unfunded liability (Trillions) 8.98%
Inflation (December) 2.07%
SSA Wage Index 2.74%
TOTAL WAGES (billions) 3.28%
SS-OASI Trust Fund Balance (thousands) 7.13% (down from 8% in 2014 negative cash flow)
SSA COLA index 2.06%
SSA Expenses (OASI) 4.95%
SSA taxation of benefits (OASI) 6.47%
TOTAL SS expenses (billions) 5.15%
SSA Payroll Revenue 3.02%
Growth in the national debt is 8% while the growth in GDP is 3.5% (2000 to 2016)
Growth in wages is 3.3% while growth rates in National Debt, Unfunded liability and SS expenses are far higher.
Growth in SS expenses is 5.15% while growth in SS tax revenues is 3.02%. Growth in the revenue from taxation of benefits is 6.47%, but will flatten as nearly all all benefits are subjected to taxation.
References:
CPI https://data.bls.gov/pdq/SurveyOutputServlet
ssa WAGE https://www.ssa.gov/oact/cola/AWI.html
SSA COLA https://www.ssa.gov/oact/cola/colaseries.html
NATIONAL DEBT https://treasurydirect.gov/govt/reports/pd/histdebt/histdebt.htm
OASI TRUST FUND https://www.ssa.gov/oact/STATS/table4a1.html
CPI UN ADJUSTED https://inflationdata.com/Inflation/Consumer_Price_Index/HistoricalCPI.aspx?reloaded=true
I know you guys don't like that Social Security is Pay-As-You-GO, but since it is, that is how you should analyze it. The unfunded liability is not useful - cash flow is the issue. There is a problem and it grows with time, but unfunded liability would grow forever even if cash flow were in balance forever.
Using infinite horizon just makes the meaningless numbers bigger.
Arne posted
"The unfunded liability is not useful - cash flow is the issue. There is a problem and it grows with time, but unfunded liability would grow forever even if cash flow were in balance forever."
An Actuarial is the correct approach. As you stated, a cash flow analysis shows the unfunded liability would grow forever. Theoretically speaking the last person to be alive would have no cash flow because there would be no one to pay their SS benefit.
The Actuarial puts the problem into perspective. The payroll tax is 10.035%. The CBO and SSA say this is not enough. Evidence shows that since 2010, it has not been enough. This means we know that benefits need to be cut or taxes raised. What if we knew or had a good idea of the cost in terms of payroll tax in 2060? Would that help people today make the decision to keep Social Security or to scrap it?
I believe that if workers knew that the cost in terms of payroll tax would have to reach 18.5% to 19.5% with a birth rate of 2.1 babies per woman, that most would say Social Security was not worth it. The simple reason is that if you put that 18.5% to 19.5% of your wages into paying down your house, retirement savings, healthcare, debt, that your net worth would be far higher and produce a substantially larger yearly income by twice that of Social Security invested in assets making only minimum returns, not high returns.
I agree that the cash flow will always produce a growing unfunded liability - the reason is simple the root cause has never been addressed. The 75 year solvency period is a sham. Comparing the cost in terms of GDP is a sham. Leaving out the majority of the liability that accrues outside the 75 year solvency period, but accrued within the 75 year solvency period is fraud.
"payroll tax would have to reach 18.5% to 19.5% with a birth rate of 2.1 "
None of these numbers is consistent with what others say. Does that make William a fraud?
Whether we blow up in a nuclear war, are invaded by little green men, or Christ comes again, the unfunded liability does not need to be paid. The root cause of that fact is that SS is PAYGO and it is the ratio of worker to beneficiaries which determines how large the benefits can be and maintain balanced cash flow.
That I believe that infinite horizon is stupid analysis does not mean that I think everyone should depend on SS for all of their retirement. A person who actually works 40 years and saves a second 10.8 percent of their income will have a solid expectation of getting more income during their expected lifetime from their personal account than from SS. But SS is insurance. It provides a safety net to people who end up not working 40 years, or end up living an extra decade.
"it is the ratio of worker to beneficiaries which determines how large the benefits can be and maintain balanced cash flow."
This means one can easily expect the payroll tax to reach 18%+. Is this a good deal for workers, no?
Arne says that " SS is insurance. It provides a safety net to people who end up not working 40 years, or end up living an extra decade."
If SS is insurance, than why does everyone who turns 65 and has paid ten years receive SS? What is the loss that SS insures against? There are few requirements to receiving this insurance:
Be a covered beneficiary with a minimum of 40 quarters of qualified work history.
Be the dependent of a worker who has qualified for SS.
Be at least age 62 and one month.
And a host of other lesser known alternatives.
Yet none are based on income or net worth. Yet insurance requires a contract stating a premium in return for covered loss(s). SS-OASI does identify a loss? Living long enough to collect is not a loss.
Insurance was added to the Social Security Act in order to not have the supreme court rule it unconstitutional.
Post a Comment