Steve Johnson post on 1988 vs 2015

[National Dialogue Moderator <moderator>]

    Mr. Johnson raises an important point with respect to the relationship
between dollars contributed to Social Security and measurements of the quality
of the investment "deal" that the system provides.

    Mr. Johnson's inquiries explore an area of policy that I believe to be
inadequately discussed, and that is the inadequacy of these standard
measurements of measuring the Social Security "rate of return."

    As Mr. Johnson correctly implies, measurements of rate of return can be
distorted somewhat if we simply track an individual's contributions to the
system through the payroll tax, in comparison to the benefits that are received
later.  The interrelationship between overall benefits and all taxation is
critical.  For example, if we abolished the payroll tax today, and funded the
system wholly through general revenues, rates of return would be determined to
be infinite (no taxes, but significant benefits.)  The President's proposal to
move to general revenue financing carries a significant form of this measurement
problem.  For a system that is largely financed through general revenues, it
would no longer be meaningful to examine the rate of return solely be comparing
payroll tax contributions and benefits.

    Similarly, a system that involved a large transition involving new debt
would include debt and interest costs that would have to be measured.  The mix
of general revenue outlays, payroll tax burdens, required savings contributions,
and debt servicing burdens are all important components of measuring the
treatment of individuals by Social Security. 

    Mr. Johnson's questions create an opportunity to describe the particular
distortions of current practices of measuring rates of return.  In the 1980s and
1990s, for example, Social Security taxes have been in surplus. Thus, they have
been used effectively to reduce the other borrowing that the federal government
must do, and this has indirectly subsidized the operations of government that
are financed through income taxation, reducing pressure on income taxes.

    In 2020, by contrast, the situation would be exactly the opposite.  Then
general revenues must make substantial outlays to Social Security.  The Social
Security system will under current law be in severe deficit, the gap made up by
general income taxation of workers at that time.

    Consequently, a worker who finished his working career in 1995 and one who
labors in the years 2015 and onward would experience very different effects. 
The former would have had his income taxes indirectly reduced by the collection
of excess payroll taxes.  Individuals working in the latter years would see
large increases in general tax payments, due to the necessity of making up the
annual shortfalls in the Social Security system.  Even though the payroll tax
rates might be the same in the two cases, the overall tax burdens are extremely
different.

    In sum, the total analysis of the system's equity cannot ignore the impact
of the Social Security cash flow upon general revenues.  These impacts are
enormous.  Some generations work at a time when approximately 2% of payroll from
Social Security subsidizes the non-Social Security budget, whereas others will
work at a time when roughly 5% of payroll must be transferred in the other
direction.  Workers will be greatly affected by where they happen to be in the
timeline when these changes occur.

    This is one reason why our proposal involved so much careful study of the
relationship between the Social Security system and annual cash flows.  This is
an important element of reform.  It is inadequate to simply propose a reform
that appears to generate "actuarial solvency" but which leaves these issues of
equity unaddressed.

    On the last question of whether a dollar contributed in 1988 and a dollar
contributed in 2015 should each be counted as a tax burden, my answer would be
contingent upon whether that dollar was saved.  Under our current structure, a
dollar contributed in 1988 was simply spent.  It was used to offset the other
borrowing that underwrote government consumption.  If an individual puts a
dollar away in 1988 and then redeems it in 2015 to receive retirement benefits,
one could make an argument that only the first dollar should be counted as the
cost of retirement savings.  If, however, an individual simply declared that a
dollar in 1988 was to be allocated for retirement, but instead spent that dollar
on a cheeseburger, promising to pay himself back in 2015, then the dollar would
have to be newly sought in 2015, and would have to be counted a second time. 
This is why the tax burdens in 1988 and in 2015 both matter under the current
system, because neither dollar is saved, and benefits are paid purely and simply
by taxing current workers, regardless of the intragovernmental accounting that
accompanies the transactions.  The intragovernmental accounting of the levels in
the Trust Fund should not disguise the reality that we are simply taxing as is
necessary to pay benefits first in 1988, and then at higher levels in 2015.  In
a pay as you go system, it is truly as simple as that, Trust Fund balances
notwithstanding.

    Senator Judd Gregg