ssa models

[National Dialogue Moderator <moderator>]

	Re: Ridgeway, "Where Do We Go From Here"

One of the great frustrations of people who study social security
and attempt to develop reform ideas--and get ready to join this
group of frustrated people--is the general unavailability of models
that generate results that are viewed as worthwhile or legitimate
in Washington.  Policymakers rely on basically one source:  the
model(s) developed by the Social Security Administration's Office
of the Actuary.  Other models have been developed by well-respected
economists (including Harvard University's Martin Feldstein and
Boston University's Laurence Kotlikoff) and they have generated
some interesting results.  But "when the rubber hits the road,"
all anybody wants to see is "the OACT memo."  (You might have picked
up on this in the other forum on reform.)

This presents many problems, big and small, for people attempting
to acquire useful information and/or develop sound public policy.
Policymakers and analysts--not to mention the public--rely on the
output of literally a handful of people who work in the government
agency administering the programs we seek to reform.  (The work of
the actuaries is generally regarded to be of high quality, but
there is only so much they can do--and only so much knowledge they
can possess--and all of their work must pass muster before political
administrators as well as political officials outside of SSA who
help establish major assumptions.)  The result is a far narrower
range of issues explored and of options developed, far fewer
estimates prepared, and a far narrower concept of the range of
possible outcomes than we would like and policymakers should have.
Here, as in virtually all other realms of activity, true competition
would be enormously beneficial.

Just think about it.  Wouldn't it be great to be able to calculate
or quickly gain access to information on social security's unfunded
liability and decide which of the many ways you'd like to use to
define "accrued benefits?" Wouldn't it be great to be able to see
how much of that liability was on account of people in different
age groups so you could think about different transition plans for
reform?  Wouldn't it be great to see lots of different ideas--all
properly "costed out"--about how a voluntary system of personal
accounts might be structured?  Wouldn't it be interesting to see
the results of models that employed SSAs data and possibly its
actuarial assumptions but incorporated their own economic assumptions?
(SSA assumes that all reforms, ranging from the 2% tax increase
"solution" to full-scale privatization have the same effect on
economic activity--none.)

To date, SSA has been unwilling to make its model(s) generally
available to the public, even with all the specs, qualifications,
and disclaimers that may be necessary or appropriate.  The public
loses out and so does SSA, which loses the benefit of learning from
the comments that would be forthcoming from the people who use and
evaluate their models.

More on this topic in my supplemental statement entitled "Assumptions
and Presumptions," in the Final Report of the 1994-1996 Social
Security Advisory Council.


Carolyn Weaver