Q: Would prolonged economic growth save Social Security?
A: No, Congress will have to face the hard choices of higher taxes or lower
benefits sooner or later. The Washington Times, 10-25-99
A booming stock market during the last eight years has
inspired the notion that steady, prolonged growth of the gross domestic product
virtually will save the Social Security trust fund from insolvency by the year
2034. Skepticism is in order for several reasons. First, it is too early to tell
if we are entering a long period of robust growth. A spell of good weather is
not necessarily a permanent climate shift. Should the economy still be expanding
without interruption five years from now, the case for optimism will be stronger.
However, nobody has repealed the business cycle. Recessions do happen. There was
much talk of a "new era" of permanent prosperity in early 1929, too.
Second, it is doubtful that the brisk growth that could help Social Security
can be achieved.
Obviously, growth depends heavily on productivity (output
per man- hour or per worker). The latest report of Social Security's Board of
Trustees includes an "intermediate" actuarial analysis in which Social
Security's trust fund is exhausted in 2034 and a "low-cost" (optimistic)
one in which trust-fund exhaustion is avoided. The productivity growth assumption
underlying the low-cost outcome is 1.6 percent per year. Put another way, this
is roughly the productivity growth needed to avert disaster in Social Security.
Productivity growth for 1958 to 1997 averaged 1.7 percent annually, which
makes the optimistic assumption look plausible. But this seemingly healthy figure
masks a steady decline, revealed by disaggregating 1958 to 1997 into 10-year periods:
2.9 percent average annual productivity growth from 1958 to 1967, 2 percent from
1968 to 1977, 1 percent from 1978 to 1987 and just 0.9 percent from 1988 to 1997.
So, for productivity growth to reach 1.6 percent, it has to nearly double
from its 1988-1997 level - and to avert Social Security's ruin, it must stay there
long term. But how likely is this to happen? Productivity is determined primarily
by investment, and the seed corn of investment is savings. Our net national savings
rate - net private savings minus federal-budget deficits, i.e., the share of gross
domestic product, or GDP, actually available for investment - has collapsed from
an average of 9.2 percent of GDP in the sixties to just 2.6 percent in the period
of 1991 to 1996. Reflecting the scarcity of investable capital, the real interest
rate - as measured by the interest rate on Moody's AAA-rated bonds minus the Consumer
Price Index, which gives a measure of the cost of credit for investment in plant
and equipment - has been steadily rising; its 1987-1996 average of 4.8 percent
is almost triple its 1957-1966 level of 1.8 percent. Just where will the money
for the investment needed for sustained doubling of productivity growth -
and robust output growth - come from? Overseas? Don't count on it. The Japanese
are struggling out of financial meltdown. They can't do our saving for us any
more.
The productivity growth slump also reflects the shift from manufacturing to services.
In 1950, 47.2 percent of the private nonagricultural labor force was in manufacturing,
52.8 percent in services. In 1995, 24.7 percent was producing goods, 75.2 percent
producing services. And while manufacturing has seen impressive productivity gains,
because its share of total output has fallen, so has its contribution to overall
productivity growth, which has been less than that in manufacturing, implying
unimpressive productivity growth in services.
True, measuring productivity
in services is difficult, and some observers argue that service-sector productivity
has been soaring since 1995. Perhaps - but again, beware of mistaking possibly
only temporary improvement for a permanent trend; whether substantial productivity
improvement sustainable over the long run indeed has occurred will not be clear
for several years yet.
Also, much of the purported productivity surge
has been achieved by replacing managerial and clerical employees with computers.
But we must be clear about what is going on here. When a labor-saving device is
introduced, productivity grows while labor savings are being realized - but once
all the savings that can be captured have been captured, productivity, though
now at a higher level, stops growing unless further enhanced by some other cause.
Just how much more blood can be wrung from this turnip? Firms have already
squeezed their workforces rigorously. Indeed, today's high labor demand indicates
that substitution of computers for labor has been taken about as far as it can
be.
Finally, remember that productivity comes down to how quickly the physical process
of producing a good or service is completed in a unit of time. Productivity growth
means that this process is being completed faster. But can output per man-hour
really rise indefinitely? This implies that eventually people will produce almost
instantaneously. Since this is obviously impossible, it follows that there are
limits on productivity growth.
Annual productivity growth of 1.6 percent
implies that in 2044 output per man-hour will be twice as high as it is now. Is
it seriously possible that in 45 years workers will build cars twice as quickly,
barbers will cut twice as many heads of hair per hour and keyboard operators will
perform twice as many keystrokes a minute? Even if we had the savings to finance
the needed investment - which we don't - it is more likely that productivity growth
eventually will taper off, and productivity will stagnate.
Given all
this, even the "intermediate" assumption of annual productivity growth
of 1.3 percent looks unlikely.
The dependence of growth on net national
savings and productivity growth is borne out by the fact that while these collapsed,
so did average annual real GDP growth, from 4.4 percent in the sixties to 2.8
percent in the eighties and just 2.2 percent in the booming nineties. And even
with the assumed rates of productivity growth, annual GDP growth for the next
10 years under the actuaries' intermediate assumptions is 2.0 percent, and under
optimistic assumptions, 2.4 percent. Given that annual GDP growth from 1960 to
1998 averaged 3.2 percent, this is underwhelming.
We could try to offset
our dismal savings performance and achieve faster economic growth by adding lots
of workers through immigration. But the foregoing unimpressive growth outlooks
assume annual immigration levels of 1.15 million (low cost) and 900,000 (intermediate)
- roughly similar to current levels. Pursuing faster growth through immigration
would entail much higher numbers - and spark a political crisis, since immigration
already is an explosive issue.
The idea that shuffling information among
computers will propel a "new economy" of fabulous growth is far-fetched.
Most production, after all, is and always will be about goods and noninformation
services - precisely where the troubling savings and productivity trends matter.
Yes, sustained brisk growth may occur. But it's unlikely, and pinning our
hopes on it is like children in a fairy tale assuming that a good fairy will appear
in time to save them from the big bad wolf. And, if she doesn't?
Third,
even if the economic growth occurs that the "new economy" enthusiasts
predict, they forget that the aging of our population will put federal finances
under severe pressure from several other sources besides Social Security.
Under its latest intermediate actuarial assumptions, which factor in recent economic
growth, assume steady growth of 2 percent through 2007 and no recessions, Medicare's
Hospital Insurance (HI, or Medicare A) "trust fund" is projected to
run deficits starting in 2007 and be exhausted in 2015. HI spending is projected
to soar to $335 billion in 2015 and $908 billion in 2030, with cash deficits for
these years of $46 billion and $330 billion, respectively. And spending for Medicare's
Supplementary Medical Insurance (SMI, or Medicare B), which pays for outpatient
hospital care, doctor visits and so on for old and disabled persons, is already
growing faster than the economy and is projected to keep doing so. In calendar-year
1998, SMI spent $77.6 billion, about one percent of GDP. Under SMI's intermediate
assumptions, outlays are projected to more than double by 2008, to $174.2 billion
- this even before the baby boomers start retiring - and reach about 2.5 percent
of GDP 30 years from now.
Spending for Medicaid, the health-care program
for the poor, will explode, too. Medicaid is the primary source of public funding
for long-term care for the elderly, financing about 31 percent of it. In 1995
Medicaid spent $28.5 billion on long-term care. While persons age 65 and older
are 13 percent of America's population now, they will be roughly 20 percent of
the population in 2030. Moreover, the population of the oldest old (aged 85 and
over), those most likely to need long-term care, will more than double by 2030,
from 3.9 million in 1997 to about 8.5 million, and more than double again by 2050,
to roughly 18 million. The budgetary implications are obvious.
Meanwhile,
as of 1996 the civil-service retirement system had unfunded liabilities of a present
value of $1.1 trillion, the military retirement system had an unfunded liability
with present value of more than $713.4 billion and the veterans' benefits program
had an unfunded liability with present value of more than $190 billion - a total
charge on future output of more than $2 trillion.
Given these other simultaneous
large fiscal pressures, growth is unlikely to help Social Security. Much of the
revenue from it will be devoured by these other needs.
Finally, the stakes
are too high for trusting in uncertain reeds such as economic growth. Since
virtually all Americans participate in Social Security, the political ramifications
of a Social Security crisis are mind-boggling. Outside help such as growth will
be nice if we can get it, but the responsible course is to refuse to lean on luck
and to grapple with overhauling the program itself.
Attarian,
an economist based in Ann Arbor, Mich., writes frequently on economic issues for
the National Review, Reason and the Detroit News.
