Reforming Social Security the Right Way

Social Security (and specifically the Old Age, Survivors and Disability Insurance, or OASDI) is in deficit.  We are collecting less in Social Security taxes than we pay out in benefits,  the trust fund is declining and is expected to be fully depleted in sixteen years, in 2035. Baby Boomers are retiring in record numbers, the number of workers per beneficiary has fallen below three, and by 2035 we expect 2.3 workers per beneficiary.  That’s quite a potential burden facing workers in the future.
In light of these facts that have been staring us in the face for decades, what proposals would one expect from serious policymakers?  Perhaps a certain combination of cutting benefits over time and some increase in taxes.
Rather than use such a balanced approach, recent proposals under consideration in Washington would raise both benefits and taxes! These proposals prominently feature provisions that would increase Social Security benefits, making the program even more costly to finance, and importantly, they would expand rather than constrain the growth of the federal government.[1]
One proposed feature in recent proposals is to increase the "replacement" factor for the lowest portion of average monthly earnings, essentially increasing Social Security benefits to lower earners. This is despite the fact that we already have a means-tested program called Supplemental Social Security to provide relief to low income retirees.  Another proposed feature is to make the annual cost of living adjustment more generous, by choosing a price index that grows faster than the current Consumer Price Index for Urban Wage Earners and Clerical Workers (CPI-W).  This would increase Social Security payments across the income distribution.  While serious academic research suggests that the CPI-W overstates inflation, we have politicians asking to index Social Security to price indices that generate even higher increases in payouts. 
These increases in benefits to current and future retirees seem misguided as a matter of policy, but certainly could garner support and, most importantly, votes.  Retirees vote, and yet with the current proposals the cost of the additional benefits would be borne by higher earning workers in future years.  Expansion of benefits for current and near-term retirees – made up of members of the Baby Boom generation – would be ironic. They are the reason the worker to beneficiary ratio in the past was high (and their tax rates were low) and they are also the reason the worker to beneficiary ratio in coming years will decline. So, the children and grandchildren of the Baby Boom generation would have to pay higher taxes to finance the Social Security benefits of the Boomers.
The current crop of reform proposals address revenue needs as well. As a point of reference, in its current form the Social Security funding gap requires a 23 percent hike in taxes, a jump from 12.4% of payroll to 15.24% of payroll, assuming no change in the benefit structure and assuming notional Trust Fund accounting.  However, that will not be enough to fund the current proposals before Congress that further expand benefits.  Importantly, these proposals would also raise taxes on higher income workers.  The revenue grab is essentially another ‘tax the rich’ scheme in the guise of Social Security reform.
The current combined OASDI payroll tax rate of 12.4% makes workers’ labor effort less valuable to workers than it is to firms and society.  This ‘tax wedge’ distorts labor supply decisions.  Raising the payroll tax rate from the current 12.4% to 15.24% would cause additional efficiency losses, and the efficiency losses will likely be even larger if the additional tax revenues were instead raised through applying the current rate to earnings above and beyond the taxable maximum currently in place, $132,900 in 2019 (and indexed to wage increases).  The taxable maximum works to render the Social Security payroll tax equivalent to a non-distortionary lump-sum tax for workers with earnings at or above the taxable maximum. Removing the taxable maximum is equivalent to a sudden marginal tax rate increase of 12.4% at existing rates, or 15.24% at the proposed rate, on high-income workers.  This is hardly a marginal change in tax rates!
Social Security’s existing funding gap is causing an increase in the federal government’s debt.  Further, Social Security benefits that have accrued to current participants is another form of government debt, basically equivalent to the better-known publicly held debt.  Social Security, by spending workers’ contributions on current retirees and then promising workers future retirement benefits, is functionally the same as a government issuing bonds equal to the accrued benefits. The implicit Social Security debt, measured by the present value of future costs for current participants less the present value of future dedicated tax incomes for current participants, was estimated to be $35.3 trillion as of January 1, 2018.  In comparison, the federal government debt – consisting of publicly held debt and intergovernmental holdings, but not including the implicit Social Security debt – was $21.5 trillion as of September 30, 2018, and U.S. GDP was $20.5 trillion in 2018.  Like the publicly held federal debt, the Social Security debt crowds out capital investment and impedes economic growth.

We have saddled our children, and our grandchildren, with a large explicit government debt and an even larger implicit or hidden government debt.  Is this to be the legacies of “The Greatest Generation” and the “Baby Boom Generation?”  Shouldn’t we ask if we have been, and are being, reasonable to our children and our grandchildren?  Have we provided them assets of equal value?  Or have we been running up a debt to fund our own current consumption activities, and essentially forcing our children to pay for our expenditures?
Sensible Social Security reform should focus on restoring long-term budget balance in OASDI, not on expanding program benefits.  Various options to gradually reduce benefits should be considered before resorting to tax increases, as reducing Social Security benefits helps reduce accrued benefits payable to current Social Security participants.  There are two main benefit-cutting reform options: (gradually) raising the retirement age and (gradually) making the benefit formula less generous for middle- and high-income workers. Note that the latter change affects higher income workers, but unlike a tax increase on these workers, it constrains, rather than increases, the size of the program. 
Social Security’s normal retirement age (or NRA) has been incrementally increasing according to a schedule adopted in the last major ‘budget-balancing’ Social Security reform in 1983.  That was 36 years past, and since then we have seen further increases in life expectancy as well as a decrease in the fertility rate, further reducing the number of workers supporting each retiree.  We could gradually raise the NRA to age 69 and then index it to life expectancy, to maintain a constant ratio of expected years in retirement to potential work years in order to stabilize the declining ratio of workers to retirees.
Social Security already redistributes income from high earners to lower earners by having three different “replacement” factors.  These are 90% for average monthly earnings smaller than $926, 32% for average monthly earnings between $926 and $5,583, and 15% for average monthly earnings larger than $5,583.  (The $926 and $5,583 values are 2019 “bend points” which are adjusted annually according to the Average Wage Index).  Although this benefit formula is already progressive, Social Security benefits might be gradually reduced for middle- and higher-income workers by making the formula even more progressive. For example, the second factor might be (gradually) reduced from 32% to 30%, a third factor from 15% to 10% and a final factor of 5% could be added at higher earnings levels.  
The Social Security actuaries have estimated that raising the normal retirement age to 69 and indexing it thereafter to mortality gains would close 39% of the current actuarial deficit all by itself.  Making the benefit formula more progressive as outlined above would reduce the deficit by 33% all by itself. And by itself raising the taxable maximum to ensure that 90% of all taxable payroll is taxed would reduce the deficit by 28%. While not strictly additive, when done altogether, these three reforms would essentially close the actuarial deficit.[2]
Importantly, the necessary changes should start sooner rather than later, so as to more evenly spread across generations the burden of fixing the current and expected future Social Security deficits.  Waiting has not, and will not, diminish the magnitude of this task.   
[1]The Social Security Actuaries have estimated the additional expenditures and revenues under several recent proposals. See the actuaries’ estimates of the “Social Security 2100 Act” and the “Social Security Expansion Act” at
[2] See the Social Security actuaries estimates of each of the provisions at

Posted: June 04, 2019 by Dennis W. Jansen, Liqun Liu, Andrew J. Rettenmaier