Is the United States Bankrupt?

Laurence J. Kotlikoff

Is the United States bankrupt? Many would scoff at this notion. Others would argue that financial
implosion is just around the corner. This paper explores these views from both partial and general
equilibrium perspectives. It concludes that countries can go broke, that the United States is going
broke, that remaining open to foreign investment can help stave off bankruptcy, but that radical
reform of U.S. fiscal institutions is essential to secure the nation’s economic future. The paper
offers three policies to eliminate the nation’s enormous fiscal gap and avert bankruptcy: a retail
sales tax, personalized Social Security, and a globally budgeted universal healthcare system.

Federal Reserve Bank of St. Louis Review, July/August 2006, 88(4), pp. 235-49.

I
s the U.S. bankrupt? Or to paraphrase the the financial markets have a long and impressive
Oxford English Dictionary, is the United record of mispricing securities; and that financial
States at the end of its resources, exhausted, implosion is just around the corner.
stripped bear, destitute, bereft, wanting in This paper explores these views from both
property, or wrecked in consequence of failure partial and general equilibrium perspectives. The
to pay its creditors? second section begins with a simple two-period
life-cycle model to explicate the economic mean-
Many would scoff at this notion. They’d point
ing of national bankruptcy and to clarify why
out that the country has never defaulted on its
government debt per se bears no connection to a
debt; that its debt-to-GDP (gross domestic product) country’s fiscal condition. The third section turns
ratio is substantially lower than that of Japan and to economic measures of national insolvency,
other developed countries; that its long-term namely, measures of the fiscal gap and genera-
nominal interest rates are historically low; that tional imbalance. This partial-equilibrium analy-
the dollar is the world’s reserve currency; and sis strongly suggests that the U.S. government is,
that China, Japan, and other countries have an indeed, bankrupt, insofar as it will be unable to
insatiable demand for U.S. Treasuries. pay its creditors, who, in this context, are current
Others would argue that the official debt and future generations to whom it has explicitly
reflects nomenclature, not fiscal fundamentals; or implicitly promised future net payments of
various kinds.
that the sum total of official and unofficial liabili-
The world, of course, is full of uncertainty.
ties is massive; that federal discretionary spending
The fourth section considers how uncertainty
and medical expenditures are exploding; that the changes one’s perspective on national insolvency
United States has a history of defaulting on its and methods of measuring a country’s long-term
official debt via inflation; that the government has fiscal condition. The fifth section asks whether
cut taxes well below the bone; that countries hold- immigration or productivity improvements aris-
ing U.S. bonds can sell them in a nanosecond; that ing either from technological progress or capital

Laurence J. Kotlikoff is a professor of economics at Boston University and a research associate at the National Bureau of Economic Research.

© 2006, The Federal Reserve Bank of St. Louis. Articles may be reprinted, reproduced, published, distributed, displayed, and transmitted in
their entirety if copyright notice, author name(s), and full citation are included. Abstracts, synopses, and other derivative works may be made
only with prior written permission of the Federal Reserve Bank of St. Louis.

F E D E R A L R E S E R V E B A N K O F S T. LO U I S R E V I E W J U LY / A U G U S T 2006 235
Kotlikoff

deepening can ameliorate the U.S. fiscal condition. an open or closed economy. Corn can be either
While immigration shows little promise, produc- consumed or used as capital (planted to produce
tivity improvements can help, provided the gov- more corn). Agents work full time when young
ernment uses higher productivity growth as an and consume when old. There is no change over
opportunity to outgrow its fiscal problems rather time in either population or technology. The popu-
than perpetuate them by effectively indexing lation of each cohort is normalized to 1.
expenditure levels to the level of productivity. Let wt stand for the wage earned when young
We certainly have seen major changes in by the generation born in year t, rt for the return
technology in recent decades, and these changes on capital at time t, and ht for the amount the
have coincided with major increases in measured government receives from the young and hands
productivity. But whether or not technology will to the old at time t.
continue to advance is an open question. There is, The generation born at time t maximizes its
however, a second source of productivity improve- consumption when old, ct + 1, subject to
ments, namely, a rise in capital per worker (capital
ct +1 h
deepening), to consider. The developed world is (1) ≤ w t − ht + t +1 .
not saving enough and will not be saving enough 1 + rt +1 1 + rt +1
to generate capital deepening on its own. However, If the economy of this country, called
China is saving and growing at such extraordi- Country X, is open and agents are free to borrow,
narily high rates that it can potentially supply the ht can exceed wt. However, consumption can’t be
United States, the European Union, and Japan negative, hence,
with huge quantities of capital. This message is
delivered in Fehr, Jokisch, and Kotlikoff (2005), ht +1
(2) ht − ≤ wt .
which simulates the dynamic transition path of 1 + rt +1
the United States, Japan, the European Union, and
China. Their model suggests that China can serve The left-hand side of (2) is generation t’s remain-
as America’s saver and, consequently, savior, ing (in this case, entire) lifetime fiscal burden—
provided the U.S. government lets growth outpace its generational account. Equation (2) says that
its spending and provided China is permitted to the government can’t extract more from a genera-
invest massive sums in our country. Unfortunately, tion than its lifetime resources, which, in this
recent experience suggests just the opposite. model, consists simply of lifetime earnings.
The final section offers three radical policies Suppose that, to keep things simple, the
to eliminate the nation’s enormous fiscal gap and economy is small and open and that the wage
avert bankruptcy. These policies would replace and interest rates are positive constants equal to
the current tax system with a retail sales tax, w and r, respectively. Also suppose that starting
personalize Social Security, and move to a glob- at some time, say 0, the government announces a
ally budgeted universal healthcare system imple- policy of setting ht equal to h forever and that
mented via individual-specific health-insurance h
vouchers. The radical stance of these proposals (3) ht − > w,
1+ r
reflects the critical nature of our time. Unless the
meaning that the generational accounts of all
United States moves quickly to fundamentally
generations starting with the one born at time 0
change and restrain its fiscal behavior, its bank-
exceed their lifetime resources.
ruptcy will become a foregone conclusion.
The old at time 0 have a generational account
(remaining lifetime fiscal burden) of –h. These
FISCAL INSOLVENCY IN A oldsters, who may have voted for the government
based on the promise of receiving h, represent the
TWO-PERIOD LIFE-CYCLE MODEL creditors in this context. But the government can’t
Consider a model in which a single good— deliver on its promise. The young may be fanati-
corn—is produced with labor and capital in either cally devoted to the government, worship the

236 J U LY / A U G U S T 2006 F E D E R A L R E S E R V E B A N K O F S T . LO U I S R E V I E W
 Kotlikoff

elderly, and care little for themselves, but they wt, the inability to borrow abroad will plunge the
cannot beg, borrow, or steal this much corn to give country into bankruptcy, assuming the government
to the government. The government can go hat in sets h as high as possible. But bankruptcy may
hand to foreign lenders, but to no avail. Foreign arise over time even if h is set below w0. To see
lenders will realize the government won’t be able this, note that capital per worker at time 1, k1, will
to repay. equal w0 – h. If w(k1) < w0, where w(kt ) (with
The most the government can do for the elderly w ′( ) > 0) references the wages of generation t,
is to set h equal to (1 + r)w/r. Let’s assume the the country will find itself in a death spiral for
government does this. In this case, the government sufficiently high values of h or sufficiently low
impoverishes each generation of young from values of w0.1 Each period’s capital stock will be
time 0 onward in order to satisfy the claims of smaller than the previous period’s until t*, where
time-0 oldsters. In the words of the Oxford English h $ wt*, making kt*+1 = 0, at which point the jig
Dictionary, we have a country at the end of its is up, assuming capital as well as labor is required
resources. It’s exhausted, stripped bear, destitute, to produce output.
bereft, wanting in property, and wrecked (at least In short, general equilibrium matters. A policy
in terms of its consumption and borrowing capac- that looks sustainable based on current conditions
ity) in consequence of failure to pay its creditors. may drive a country broke and do so on a perma-
In short, the country is bankrupt and is forced to nent basis. Of course, policymakers may adjust
reorganize its operations by paying its creditors their policies as they see their country’s output
(the oldsters) less than they were promised. decline. But they may adjust too little or too late
and either continue to lose ground or stabilize
Facing the Music their economies at very unpleasant steady states.
The point at which a country goes bankrupt Think of Argentina, which has existed in a state
depends, in general, on its technology and prefer- of actual or near-bankruptcy for well neigh a
ences as well as its openness to international trade. century. Argentina remains in this sorry state for
If, for example, agents who face confiscatory life- a good reason. Its creditors—primarily each suc-
time fiscal burdens refuse to work, there will be cessive generation of elderly citizens—force the
no lifetime resources for the government to appro- government to retain precisely those policies that
priate. Consequently, the government must further perpetuate the country’s destitution.
limit what it can pay its creditors.
As a second example, consider what happens Does Official Debt Record or Presage
when an open economy, which has been transfer- National Bankruptcy?
ring (1 + r)w/r to the elderly on an ongoing basis, Since general equilibrium considerations
suddenly, at time 0, becomes closed to interna- play a potentially critical role in assessing policy
tional trade and credit. In this case, the govern- sustainability and the likelihood of national bank-
ment can no longer pay the contemporaneous ruptcy, one would expect governments to be hard
elderly the present value of the resources of all at work developing such models or, at a minimum,
current and future workers. Instead, the most it doing generational accounting to see the potential
can pay the time-t elderly is the current young’s burden facing current young and future genera-
resources, namely wt. The reason is simple. The
tions. That’s not the case. Instead, governments
time-t young have no access to foreign loans, so
they can’t borrow against their future receipt of 1
If h is sufficiently large, there will be no steady state of the economy
h in order to hand the government more at time t featuring a positive capital stock. In this case, the economy’s capital
than wt. stock will converge to zero over time, starting from any initial
value of capital. If h is not so large as to preclude a steady state
Clearly the loss of foreign credit will require with positive capital, the economy will feature two steady states,
the government to renege on much of its commit- one stable and one unstable. The capital stock in the stable steady
state will exceed that in the unstable steady state. In this case, the
ment to the time-t oldsters. And if the government economy will experience a death spiral only if its initial capital
was initially setting h below (1 + r)w/r, but above stock is less than that in the unstable steady state.

F E D E R A L R E S E R V E B A N K O F S T . LO U I S R E V I E W J U LY / A U G U S T 2006 237
Kotlikoff

around the world rely on official debt as the government “runs” a surplus that becomes infi-
primary indicator of fiscal solvency. So do the nitely large relative to the size of the economy.
International Monetary Fund, World Bank, And because g > r, the present value of the time-t
Organisation for Economic Co-operation and surplus as t goes to infinity is infinite.
Development, and virtually all other monitors Alternatively, the government can set (use
of economic policy, including most academic words such that) mt+1 = m(1+ g), m0 = 1, and g > r.
economists. In this case, official debt becomes infinitely large
Unfortunately, the focus on government debt and the present value of government debt at time
has no more scientific basis than reading tea leaves t as t goes to infinity is infinite. So much for the
or examining entrails. To see this, let’s return to transversality condition on government debt!
our small open and entirely bankrupt Country X, Thus, the government of bankrupt Country X
which, when we left it, was setting h at the maxi- is free to say it’s running a balanced budget policy
mally expropriating value of (1 + r)w/r. Can we (by saying mt – 0 for t $ 0); a surplus policy, where
use Country X’s debt to discern its insolvency? the surplus becomes enormous relative to the size
Good question, particularly because the word of the economy; or a debt policy, where the debt
“debt” wasn’t used at all in describing Country X’s becomes enormous relative to the size of the
fiscal affairs. Neither, for that matter, were the economy. Or it could pick values of the mt s that
words “taxes” or “transfer payments.” This, by change sign from one period to the next or, if it
itself, indicates the value of “debt” as a precursor likes, on a random basis. In this case, Country X
or cursor of bankruptcy, namely, zero. But to drive would “run” deficits as well as “surpluses”
the point home, suppose Country X calls the h it through time, with no effect whatsoever on the
takes from the young each period a “tax” and the economy or the country’s underlying policy.
“h” it gives to the old each period a “transfer pay- But no one need listen to the government.
ment.” In this case, Country X never runs a deficit, Speech, or at least thought, is free. Each citizen of
never has an epsilon worth of outstanding debt, Country X, or of any other country for that matter,
and never defaults on debt. Even though it is as can choose her own language (pattern of the mt s)
broke as broke can be, Country X can hold itself and pronounce publicly or whisper to herself that
out as debt-free and a model of fiscal prudence. Country X is running whatever budgetary policy
Alternatively, let’s assume the government most strikes her fancy. Citizens schooled on
continues to call the h it gives the time-0 elderly Keynesian economics as well as supply siders,
a transfer payment, but that it calls the h it takes both of whom warm to big deficits, can choose
from the young in periods t $ 0 “borrowing of fiscal labels to find fiscal bliss. At the same time,
mt h less a transfer payment of (mt – 1)h” and the Rockefeller Republicans (are there any left and
h it gives generation t when it is old at time t + 1 do they remember Rocky?) can soothe their souls
“repayment of principal plus interest in the with reports of huge surpluses and fiscal sobriety.
amount of mt h(1+ r) less a net tax payment of To summarize, countries can go bankrupt,
–h + mt h(1+ r).” Note that no one’s generational but whether or not they are bankrupt or are going
account is affected by the choice of language. How- bankrupt can’t be discerned from their “debt”
ever, the outstanding stock of debt at the end of policies. “Debt” in economics, like distance and
each period t is now mt h. time in physics, is in the eyes (or mouth) of the
The values of mt can be anything the govern- beholder.2
ment wants them to be. In particular, the govern-
2
ment can set (use words such that) By economics, I mean neoclassical economics in which neither
agents nor economic institutions are affected by language. Kotlikoff
(4) mt +1 = mt (1 + g ), m0 = −1, and g > r . (2003) provides a longer treatment of this issue, showing that the
vapidity of conventional fiscal language is in no way mitigated by
considerations of uncertainty, time consistency, distortions, liquidity
In this case, official debt is negative; i.e., the constraints, or the voluntary nature of payments to the government.

238 J U LY / A U G U S T 2006 F E D E R A L R E S E R V E B A N K O F S T. LO U I S R E V I E W
 Kotlikoff

ECONOMIC MEASUREMENT OF Paul O’Neill. Smetters, who served as Deputy

THE U.S. FISCAL CONDITION Assistant Secretary of Economic Policy at the
Treasury between 2001 and 2002, recruited
As suggested above, the proper way to con- Gokhale, then Senior Economic Adviser to the
sider a country’s solvency is to examine the life- Federal Reserve Bank of Cleveland, to work with
time fiscal burdens facing current and future him and other Treasury staff on the study. The
generations. If these burdens exceed the resources study took close to a year to organize and complete.
of those generations, get close to doing so, or sim- Gokhale and Smetters’s $65.9 trillion fiscal-
ply get so high as to preclude their full collection, gap calculation relies on the same methodology
the country’s policy will be unsustainable and employed in the original Treasury analysis. Hence,
can constitute or lead to national bankruptcy. one can legitimately view this figure as our own
Does the United States fit this bill? No one government’s best estimate of its present-value
knows for sure, but there are strong reasons to budgetary shortfall. The $65.9 trillion gap is all
believe the United States may be going broke. the more alarming because its calculation omits
Consider, for starters, Gokhale and Smetters’s the value of contingent government liabilities
(2005) analysis of the country’s fiscal gap, which and relies on quite optimistic assumptions about
measures the present value difference between increases over time in longevity and federal
all future government expenditures, including healthcare expenditures.
servicing official debt, and all future receipts. In Take Medicare and Medicaid spending, for
calculating the fiscal gap, Gokhale and Smetters example. Gokhale and Smetters assume that the
use the federal government’s arbitrarily labeled growth rate in these programs’ benefit levels
receipts and payments. Nevertheless, their calcu- (expenditures per beneficiary at a given age) in
lation of the fiscal gap is label-free because alter- the short and medium terms will be only 1 per-
native labeling of our nation’s fiscal affairs would centage point greater than the growth rate of real
yield the same fiscal gap. Indeed, determining wages per worker. In fact, over the past four years,
the fiscal gap is part of generational accounting; real Medicare benefits per beneficiary grew at an
the fiscal gap measures the extra burden that annual rate of 3.51 percent, real Medicaid bene-
would need to be imposed on current or future fits per beneficiary grew at an annual rate of 2.36
generations, relative to current policy, to satisfy percent, and real weekly wages per worker grew
the government’s intertemporal budget constraint. at an annual rate of 0.002 percent.3
The Gokhale and Smetters measure of the Medicare and Medicaid’s benefit growth over
fiscal gap is a stunning $65.9 trillion! This figure the past four years has actually been relatively
is more than five times U.S. GDP and almost twice modest compared with that in the past. Table 1,
the size of national wealth. One way to wrap one’s taken from Hagist and Kotlikoff (forthcoming),
head around $65.9 trillion is to ask what fiscal shows real benefit levels in these programs grew
adjustments are needed to eliminate this red hole. at an annual rate of 4.61 percent between 1970
The answers are terrifying. One solution is an and 2002. This rate is significantly higher than
immediate and permanent doubling of personal that observed during the same period in Germany,
and corporate income taxes. Another is an imme- Japan, and the United Kingdom. Given the intro-
diate and permanent two-thirds cut in Social duction of the new Medicare prescription drug
Security and Medicare benefits. A third alterna- benefit, which will start paying benefits in 2006,
tive, were it feasible, would be to immediately one can expect Medicare benefit growth to
and permanently cut all federal discretionary increase substantially in the near term.
spending by 143 percent.
The Gokhale and Smetters study is an update 3
See www.cms.hhs.gov/researchers/pubs/datacompendium/2003/
of an earlier, highly detailed, and extensive U.S. 03pg4.pdf from the Centers for Medicare and Medicaid website
and http://a257.g.akamaitech.net/7/257/2422/17feb20051700/
Department of the Treasury fiscal gap analysis www.gpoaccess.gov/eop/2005/B47.xls from the 2005 Economic
commissioned in 2002 by then Treasury Secretary Report of the President.

F E D E R A L R E S E R V E B A N K O F S T. LO U I S R E V I E W J U LY / A U G U S T 2006 239
Kotlikoff

long-term fiscal problems. Senator Kerry made

Table 1 no serious proposals to reform Social Security,
Average Annual Benefit Growth Rates, Medicare, or Medicaid during the 2004 presi-
1970-2002 dential campaign. And his Democratic colleagues
in Congress have evoked Nancy Reagan’s mantra—
Country Rate (%)
“Just say no!”—in response to the president’s
Australia 3.66 repeated urging to come to grips with Social
Austria 3.72 Security’s long-term financing problem.
Canada 2.32 The Democrats, of course, had eight long years
Germany 3.30 under President Clinton to reform our nation’s
Japan 3.57 most expensive social insurance programs. Their
Norway 5.04 failure to do so and the Clinton administration’s
Spain 4.63 censorship of an Office of Management and
Sweden 2.35 Budget generational accounting study, which was
United Kingdom 3.46 slated to appear in the president’s 1994 budget,
United States 4.61
speaks volumes about the Democrats’ priorities
and their likely future leadership in dealing with
SOURCE: Hagist and Kotlikoff (forthcoming). our nation’s fiscal fiasco.
The fiscal irresponsibility of both political
parties has ominous implications for our children
How are the Bush administration and Congress and grandchildren. Leaving our $65.9 trillion bill
planning to deal with the fiscal gap? The answer, for today’s and tomorrow’s children to pay will
apparently, is to make it worse by expanding roughly double their average lifetime net tax rates
discretionary spending while taking no direct (defined as the present value of taxes paid net of
steps to raise receipts. The costs of hurricanes transfer payments received divided by the present
Katrina and Rita could easily total $200 billion value of lifetime earnings).
over the next few years. And the main goal of the Table 2, taken from Gokhale, Kotlikoff, and
President’s tax reform initiative will likely be to Sluchynsky (2003), presents the average lifetime
eliminate the alternative minimum tax. net tax rates now facing couples who are 18 years
This administration’s concern with long-term of age and work full time. The calculations incor-
fiscal policy is typified by the way it treated the porate all major tax and transfer programs and
Treasury’s original fiscal gap study. The study assume that the couples work full time through
was completed in the late fall of 2002 and was age 64, experience a 1 percent annual real earnings
slated to appear in the president’s 2003 budget growth, have children at ages 25 and 27, purchase
to be released in early February 2003. But when a house scaled to their earnings, and pay college
Secretary O’Neill was ignominiously fired on tuition scaled to their earnings. The table shows
December 6, 2002, the study was immediately that average lifetime net tax rates are already
censored. Indeed, Gokhale and Smetters were told fairly high for middle and high earners, who, of
within a few days of O’Neill’s firing that the study course, pay the vast majority of total taxes.
would not appear in the president’s budget. The The table also presents marginal net work tax
timing of these events suggests the study itself rates. These are not marginal tax rates on working
may explain O’Neill’s ouster or at least the timing full time (versus not working at all). They are
of his ouster. Publication of the study would, no not marginal net tax rates on working additional
doubt, have seriously jeopardized the passage of hours. They are computed by comparing the
the administration’s Medicare drug benefit as well present value of additional lifetime spending one
as its third tax cut. can afford from working full time each year from
For their part, the Democrats have studiously age 18 through age 64 and paying net taxes with
avoided any public discussion of the country’s the present value of additional lifetime spending

240 J U LY / A U G U S T 2006 F E D E R A L R E S E R V E B A N K O F S T. LO U I S R E V I E W
 Kotlikoff

Table 2
Average Net Full-time Worker Tax Rates
Multiple of Initial total household Average lifetime Marginal net
minimum wage earnings (2002 $) net tax rate (%) tax rate (%)
1 21,400 –32.2 66.5
1.5 32,100 14.8 80.6
2 42,800 22.9 72.2
3 64,300 30.1 63.0
4 85,700 34.4 59.1
5 107,100 37.8 57.5
6 128,500 41.0 57.5
7 150,000 42.9 57.0
8 171,400 44.2 56.6
9 192,800 45.1 56.1
10 214,200 45.7 55.7
15 321,400 48.4 55.2
20 428,500 49.6 54.7
30 642,700 50.8 54.2
40 857,000 51.4 54.0

NOTE: Present values are actuarial and assume a 5 percent real discount rate.
SOURCE: Gokhale, Kotlikoff, and Sluchynsky (2003).

one can afford in the absence of any taxes or average lifetime net tax rates of future generations
transfers. would entail layering additional highly distortive
Clearly, these marginal net tax rates are very net taxes on top of a net tax system that is already
high, ranging from 54.0 percent to 80.6 percent. highly distortive. If work and saving disincentives
The rates are highest for low-income workers. For worsen significantly for the broad middle class,
such workers, working full time can mean the we’re likely to see major supply responses of the
partial or full loss of the earned income tax credit type that have not yet arisen in this country. In
(EITC), Medicaid benefits, housing support, food addition, we could see massive emigration. That
stamps, and other sources of welfare assistance. sounds extreme, but anyone who has visited
Going to work also means paying a combined Uruguay of late would tell you otherwise. Uruguay
employer-employee Federal Insurance Contribu- has very high net tax rates and has lost upward
tion Act (FICA) tax of 15.3 percent and, typically, of 500,000 young and middle-aged workers to
state (Massachusetts, in this case) income taxes Spain and other countries in recent years. Many of
and federal income taxes (gross of EITC benefits). these émigrés have come and are still coming from
Together with David Rapson, a graduate stu- the ranks of the nation’s best educated citizens.
dent at Boston University, I am working to develop Given the reluctance of our politicians to raise
comprehensive measures of lifetime marginal net taxes, cut benefits, or even limit the growth in
taxes on working additional hours and saving benefits, the most likely scenario is that the gov-
additional dollars. Our early work suggests quite ernment will start printing money to pay its bills.
high marginal net taxes on these choices as well. This could arise in the context of the Federal
The point here is that trying to double the Reserve “being forced” to buy Treasury bills and

F E D E R A L R E S E R V E B A N K O F S T . LO U I S R E V I E W J U LY / A U G U S T 2006 241
Kotlikoff

bonds to reduce interest rates. Specifically, once permanently remains in its time-0 configuration.
the financial markets begin to understand the Further assume that w(k1) < h, so that if technology
depth and extent of the country’s financial insol- doesn’t change, the government will go bankrupt
vency, they will start worrying about inflation and in period 1.
about being paid back in watered-down dollars. How should an economist observing this
This concern will lead them to start dumping their economy at time 0 describe its prospects for
holdings of U.S. Treasuries. In so doing, they’ll bankruptcy? One way, indeed, the best way, is to
drive up interest rates, which will lead the Fed to simply repeat the above paragraph; that is, take
print money to buy up those bonds. The conse- one’s audience through (simulate) the different
quence will be more money creation—exactly possible scenarios.
what the bond traders will have come to fear. But what about generational accounting?
This could lead to spiraling expectations of How does the economist compare the lifetime
higher inflation, with the process eventuating in burden facing, for example, workers born in
hyperinflation. period 1 with their capacity to meet that burden?
Yes, this does sound like an extreme scenario Well, the burden that the government wants to
given the Fed’s supposed independence, our recent impose, regardless of the technology, entails taking
history of low inflation, and the fact that the dollar away h from generation 1 when the generation is
is the world’s principal reserve currency. But the young and giving h back to the generation when
United States has experienced high rates of infla- it’s old. Because in the regular (the non *) state
tion in the past and appears to be running the same the government will, by assumption, do its best
type of fiscal policies that engendered hyper- by its claimants (the time-1 elderly), generation
inflations in 20 countries over the past century. 1 can expect to hand over all their earnings when
young and receive nothing when old (because the
capital stock when old will be zero). This is a
INCORPORATING UNCERTAINTY 100 percent lifetime net tax rate.
The world, of course, is highly uncertain. And In the * state, the lifetime net tax rate will be
the fiscal gap/generational accounting discussed lower. Suppose it’s only 50 percent. Should one
above fails to systematically account for that then form a weighted average of the 100 percent
uncertainty. There are two types of uncertainties and 50 percent lifetime net tax rates with weights
that need to be considered in assessing a country’s equal to (1 – α ) and α , respectively? Doing so
prospects for bankruptcy. The first is uncertainty would generate a high average net tax rate, but one
in the economy’s underlying technology and below 100 percent. Reporting that generation 1
preferences. The second is uncertainty in policy. faces a high expected net tax rate conveys impor-
Let’s take the former first. Specifically, let’s tant information, namely, that the economy is
return to our two-period model but assume that nearing bankruptcy. But citing a figure less than
the economy is closed to international trade. And 100 percent may also give the false impression
let’s assume that at time 0 the economy appears that there is no absolutely fatal scenario.
to be going broke insofar as the government has Note that agents born at time 1 can’t trade in
set a permanent level of h such that the economy a market prior to period 1 in order to value their
will experience a death spiral in the absence of lifetime wages and lifetime fiscal burdens. If such
any changes in technology. Thus, k1 = w0 – h, a contingent claims market existed, there would
and w(k1) < w0, where w( ) references the wage- be market valuations of these variables (but no
generation function based on existing technology. trades because all cohort members are assumed
Now suppose there is a chance, with proba- identical). In this case, we could compare the
bility α, of the economy’s technology permanently value of claims to future earnings with the negative
changing, entailing a new and permanent wage- value of claims to future net taxes. But again, this
generation function, w*( ), such that w*(k1) > w0. comparison might fail to convey what one really
If this event doesn’t arise, assume that technology wants to say about national bankruptcy, namely,

242 J U LY / A U G U S T 2006 F E D E R A L R E S E R V E B A N K O F S T . LO U I S R E V I E W
 Kotlikoff

the chances it will occur and the policies needed government liability. The real issue is not how to
to avoid it. How about uncertainty with respect to value those liabilities, but rather who will pay
future policy? Well, the same considerations just them, assuming they end up having to be paid.
mentioned appear to apply for that case as well. The economy could operate with perfect state-
In my view, the best way for generational contingent claims markets so that we could tell
accounting to accommodate uncertainty is to precisely the market value of the government’s
establish lifetime fiscal burdens facing future contingent claims and see clearly that the govern-
generations under different scenarios about the ment’s budget constraint was satisfied—that the
evolution of the economy and of policy. This will market value of all of the government’s state-
necessarily be partial-equilibrium analysis. But contingent expenditures were fully covered by
that doesn’t mean that the projections used in the market value of its state-contingent receipts.
generational accounting have to be static and But this knowledge would not by itself tell us how
assume that neither policy nor economic vari- badly generation X would fare were state Y to
ables change through time. Instead, one should eventuate. Pricing risk doesn’t eliminate risk.
use general equilibrium models to inform and And what we really want to know is not just the
establish policy projection scenarios to which price at which, for example, the Pension Benefit
generational accounting can then be applied. Guarantee Corporation can offload its contingent
In thinking about uncertainty and this pro- liabilities, but also who will suffer and by how
posed analysis, one should bear in mind that the much when the Corporation fails to do so and
goal of long-term fiscal analysis and planning is ends up getting hit with a bill.
not to determine whether the government’s inter-
temporal budget constraint is satisfied, per se.
We know that no matter what path the economy CAN IMMIGRATION,
travels, the government’s intertemporal budget PRODUCTIVITY GROWTH,
constraint will be satisfied on an ex post basis. The OR CAPITAL DEEPENING SAVE
manner in which the budget constraint gets satis-
fied may not be pretty. But economic resources
THE DAY?
are finite, and the government must and will Many members of the public as well as offi-
ultimately make someone pay for what it spends.4 cials of the government presume that expanding
Thus, in the case of the United States, one immigration can cure what they take to be funda-
could say that there is no fiscal problem facing mentally a demographic problem. They are wrong
the United States because the government’s inter- on two counts. First, at heart, ours is not a demo-
temporal budget constraint is balanced once one graphic problem. Were there no fiscal policy in
takes into account that young and future genera- place promising, on average, $21,000 (and grow-
tions will, one way or other, collectively be forced ing!) in Social Security, Medicare, and Medicaid
to pay $65.9 trillion more than they would have benefits to each American age 65 and older, our
to pay based on current tax and transfer schedules. having a much larger share of oldsters in the
But the real issue is not whether the constraint is United States would be of little economic concern.
satisfied. The real issue is whether the path the Second, it is mistake to think that immigration
government is taking in the process of satisfying can significantly alleviate the nation’s fiscal prob-
the constraint is, to put it bluntly, morally and lem. The reality is that immigrants aren’t cheap.
economically nuts. They require public goods and services. And they
The above point bears on the question of become eligible for transfer payments. While most
valuing the government’s contingent liabilities. immigrants pay taxes, these taxes barely cover
The real economic issue with respect to contingent the extra costs they engender. This, at least, is the
liabilities is the same as that with respect to any conclusion reached by Auerbach and Oreopoulos
(2000) in a careful generational accounting analy-
4
This statement assumes that the economy is dynamically efficient. sis of this issue.

F E D E R A L R E S E R V E B A N K O F S T . LO U I S R E V I E W J U LY / A U G U S T 2006 243
Kotlikoff

A different and more realistic potential cure Assuming the United States could restrain
for our fiscal woes is productivity growth, which the growth in its expenditures in light of produc-
is supposed to (i) translate into higher wage growth tivity and real wage advances, is there a reliable
and (ii) expand tax bases and limit requisite tax source of productivity improvement to be tapped?
hikes. Let’s grant that higher rates of productivity The answer is yes, and the answer lies with China.
growth raise real average wages even though the China is currently saving over a third of its national
relationship between the two has been surpris- income and growing at spectacularly high rates.
ingly weak in recent decades. And let’s accept Even though it remains a developing country,
that higher real wages will lead to larger tax bases China is saving so much that it’s running a current
even though it could lead some workers to cut account surplus. Not only is China supplying
back on their labor supply or retire early. This capital to the rest of the world, it’s increasingly
isn’t enough to ensure that productivity growth doing so via direct investment. For example,
raises resources on net. The reason, of course, is China is investing large sums in Iran, Africa, and
that some government expenditures, like Social Eastern Europe.5
Security benefits, are explicitly indexed to pro- Although China holds close to a half trillion
ductivity and others appear to be implicitly U.S. dollars in reserves, primarily in U.S. Treasur-
indexed. ies, the United States sent a pretty strong message
Take military pay. There’s no question but that in recent months that it doesn’t welcome Chinese
a rise in general wage levels would require paying direct investment. It did so when it rejected the
commensurately higher wages to our military Chinese National Petroleum Corporation’s bid to
volunteers. Or consider Medicare benefits. A rise
purchase Unocal, a U.S. energy company. The
in wage levels can be expected to raise the quality
Chinese voluntarily withdrew their bid for the
of healthcare received by the work force, which
company. But they did so at the direct request of
will lead the elderly (or Congress on behalf of the
the White House. The question for the United
elderly) to push Medicare to provide the same.
States is whether China will tire of investing only
Were productivity growth a certain cure for the
indirectly in our country and begin to sell its
nation’s fiscal problems, the cure would already
dollar-denominated reserves. Doing so could have
have occurred. The country, after all, has experi-
spectacularly bad implications for the value of
enced substantial productivity growth in the
the dollar and the level of U.S. interest rates.
postwar period, yet its long-term fiscal condition
is worse now than at any time in the past. The Fear of Chinese investment in the United
limited ability of productivity growth to reduce States seems terribly misplaced. With a national
the implied fiscal burden on young and future gen- saving rate running at only 2.1 percent—a postwar
erations is documented in Gokhale and Smetters low—the United States desperately needs for-
(2003) under the assumption that government eigners to invest in the country. And the country
discretionary expenditures and transfer payments with the greatest potential for doing so going for-
are indexed to productivity. ward is China.6
But the past linkage of federal expenditures to Fehr, Jokisch, and Kotlikoff (2005) develop a
real incomes need not continue forever. Margaret dynamic, life-cycle, general equilibrium model
Thatcher made a clean break in that policy when to study China’s potential to influence the transi-
she moved to adjusting British government-paid tion paths of Japan, the United States, and the
pensions to prices rather than wages. Over time, European Union. Each of these countries/regions
the real level of state pensions has remained rela- is entering a period of rapid and significant aging
tively stable, while the economy has grown. As a
5
result of this and other policies, Great Britain is See www.atimes.com/atimes/China/GF04Ad07.html;
www.channel4.com/news/special-reports/
close to generational balance; that is, close to a special-reports-storypage.jsp?id=310; and
situation in which the lifetime net tax rates on http://english.people.com.cn/200409/20/eng20040920_157654.html.
future generations will be no higher than those 6
The remainder of this section draws heavily on Fehr, Jokisch, and
facing current generations. Kotlikoff (2005).

244 J U LY / A U G U S T 2006 F E D E R A L R E S E R V E B A N K O F S T . LO U I S R E V I E W
 Kotlikoff

that will require major fiscal adjustments. But the century. Without China they’d be only 2 per-
the aging of these societies may be a cloud with cent higher in 2030 and, as mentioned, 4 percent
a silver lining coming, in this case, in the form of lower at the end of the century.
capital deepening that will raise real wages. What’s more, the major outflow of the devel-
In a previous model that excluded China oped world’s capital to China predicted in the
(Fehr, Jokisch, and Kotlikoff, 2004), my coauthors short run by our model does not come at the cost
and I predicted that the tax hikes needed to pay of lower wages in the developed world. The reason
benefits along the developed world’s demographic is that the knowledge that their future wages will
transition would lead to a major capital shortage, be higher (thanks to China’s future capital accu-
reducing real wages per unit of human capital by mulation) leads our model’s workers to cut back
one-fifth over time. A recalibration of our original on their current labor supply. So the short-run
model that treats government purchases of capital outflow of capital to China is met with a commen-
goods as investment rather than current consump- surate short-run reduction in developed-world
tion suggests this concern was overstated. With labor supply, leaving the short-run ratio of physi-
government investment included, we find much cal capital to human capital, on which wages posi-
less crowding-out over the course of the century tively depend, actually somewhat higher than
and only a 4 percent long-run decline in real would otherwise be the case.
wages. One can argue both ways about the true Our model does not capture the endogenous
capital-goods content of much of government determination of skill premiums studied by
investment, so we don’t view the original findings Heckman, Lochner, and Taber (1996) or include
as wrong, just different. the product of low-skill-intensive products. Doing
Adding China to the model further alters, so could well show that trade with China, at least
indeed, dramatically alters, the model’s predic- in the short run, explains much of the relative
tions. Even though China is aging rapidly, its sav- decline in the wages of low-skilled workers in the
ing behavior, growth rate, and fiscal policies are developed world. Hence, we don’t mean to sug-
currently very different from those of developed gest here that all United States, European Union,
countries. If successive Chinese cohorts continue and Japanese workers are being helped by trade
to save like current cohorts, if the Chinese govern- with China, but rather that trade with China is,
ment can restrain growth in expenditures, and if on average, raising the wages of developed-world
Chinese technology and education levels ulti- workers and will continue to do so.
mately catch up with those of the West and Japan, The notion that China, India, and other
the model looks much brighter in the long run. developing countries will alleviate the developed-
China eventually becomes the world’s saver and, world’s demographic problems has been stressed
thereby, the developed world’s savior with respect by Siegel (2005). Our paper, although it includes
to its long-run supply of capital and long-run only one developing country—China—supports
general equilibrium prospects. And, rather than Siegel’s optimistic long-term macroeconomic
seeing the real wage per unit of human capital view. On the other hand, our findings about the
fall, the West and Japan see it rise by one-fifth by developed world’s fiscal condition remain trou-
2030 and by three-fifths by 2100. These wage bling. Even under the most favorable macroeco-
increases are over and above those associated nomic scenario, tax rates still rise dramatically
with technical progress, which we model as over time in the developed world to pay baby
increasing the human capital endowments of boomers their government-promised pension and
successive cohorts. health benefits. However, under the best-case
Even if the Chinese saving behavior (captured scenario, in which long-run wages are 65 percent
by its time-preference rate) gradually approaches higher, the U.S. payroll tax rates are roughly 40
that of Americans, developed-world real wages percent lower than they would otherwise be.
per unit of human capital are roughly 17 percent This result rests on the assumption that, while
higher in 2030 and 4 percent higher at the end of Social Security benefits are increased in light of

F E D E R A L R E S E R V E B A N K O F S T . LO U I S R E V I E W J U LY / A U G U S T 2006 245
Kotlikoff

the Chinese-investment-induced higher real wages, workers spent their wages, they would both pay
federal government healthcare benefits are not; sales taxes.
that is, the long-run reduction in payroll tax rates The single, flat-rate sales tax would pay for
is predicated on outgrowing a significant share all federal expenditures. The tax would be highly
of our healthcare-expenditure problems. transparent and efficient. It would save hundreds
of billions of dollars in tax compliance costs. And
it would either reduce or significantly reduce
FIXING OUR FISCAL effective marginal taxes facing most Americans
INSTITUTIONS7 when they work and save.
The sales tax would also enhance generational
Determining whether a country is already equity by asking rich and middle class older
bankrupt or going bankrupt is a judgment call. Americans to pay taxes when they spend their
In my view, our country has only a small window wealth. The poor elderly, living on Social Security,
to address our problems before the financial would end up better off. They would receive the
markets will do it for us. Yes, there are ways out sales tax rebate even though the purchasing power
of our fiscal morass, including Chinese investment of their Social Security benefits would remain
and somehow getting a lid on Medicare and unchanged (thanks to the automatic adjustment
Medicaid spending, but I think immediate and to the consumer price index that would raise
fundamental reform is needed to confidently their Social Security benefits to account for the
secure our children’s future. increase in the retail-price level).
The three proposals I recommend cover taxes, The sales tax would be levied on all final
Social Security, and healthcare and are intercon- consumption goods and services and would be
nected and interdependent. In particular, tax set at 33 percent—high enough to cover the costs
reform provides the funding needed to finance of this “New New Deal’s” Social Security and
Social Security and healthcare reform. It also healthcare reforms as well as meet the govern-
ensures that the rich and middle class elderly ment’s other spending needs. On a tax-inclusive
pay their fair share in resolving our fiscal gap. basis, this is a 25 percent tax rate, which is a lower
or much lower marginal rate than most workers
Tax Reform pay on their labor supply. The marginal tax on
saving under the sales tax would be zero, which
The plan here is to replace the personal is dramatically lower than the effective rate now
income tax, the corporate income tax, the pay- facing most savers.
roll (FICA) tax, and the estate and gift tax with a
federal retail sales tax plus a rebate. The rebate Social Security Reform
would be paid monthly to households, based on
My second proposed reform deals with Social
the household’s demographic composition, and
Security. I propose shutting down the retirement
would be equal to the sales taxes paid, on average,
portion of the current Social Security system at
by households at the federal poverty line with the the margin by paying in the future only those
same demographics. retirement benefits that were accrued as of the
The proposed sales tax has three highly pro- time of the reform. This means that current retirees
gressive elements. First, thanks to the rebate, poor would receive their full benefits, but current
households would pay no sales taxes in net terms. workers would receive benefits based only on
Second, the reform would eliminate the highly their covered wages prior to the date of the reform.
regressive FICA tax, which is levied only on the The retail sales tax would pay off all accrued
first $90,000 of earnings. Third, the sales tax would retirement benefits, which eventually would equal
effectively tax wealth as well as wages, because zero. The current Social Security survivor and
when the rich spent their wealth and when disability programs would remain unchanged
except that their benefits would be paid by the
7
This section draws heavily from Ferguson and Kotlikoff (2005). sales tax.

246 J U LY / A U G U S T 2006 F E D E R A L R E S E R V E B A N K O F S T . LO U I S R E V I E W
 Kotlikoff

In place of the existing Social Security retire- Healthcare Reform

ment system, I would establish the Personal
My final proposed reform deals not just with
Security System (PSS)—a system of individual
our public healthcare programs, Medicare and
accounts, but one with very different properties
Medicaid, but with our private health-insurance
from the scheme proposed by the president. All
system as well. That system, as is well known,
workers would be required to contribute 7.15
leaves some 45 million Americans uninsured. My
percent of their wages up to what is now the
reform would abolish the existing fee-for-service
earnings ceiling covered by Social Security (i.e.,
Medicare and Medicaid programs and enroll all
they’d contribute what is now the employee FICA
Americans in a universal health-insurance system
payment) into an individual PSS account. Married
called the Medical Security System (MSS). In
or legally partnered couples would share contri-
October of each year, the MSS would provide each
butions so that each spouse/partner would receive
American with an individual-specific voucher to
the same contribution to his or her account. The
be used to purchase health insurance for the fol-
government would contribute to the accounts of lowing calendar year. The size of the voucher
the unemployed and disabled. In addition, the would depend on the recipients’ expected health
government would make matching contributions expenditures over the calendar year. Thus, a 75
on a progressive basis to workers’ accounts, year old with colon cancer would receive a very
thereby helping the poor to save. large voucher, say $150,000, whereas a healthy
All PSS accounts would be private property. 30 year old might receive a $3,500 voucher.
But they would be administered and invested by The MSS would have access to all medical
the Social Security Administration in a market- records concerning each American and set the
weighted global index fund of stocks, bonds, and voucher level each year based on that information.
real-estate securities. Consequently, everyone Those concerned about privacy should rest easy.
would have the same portfolio and receive the The government already knows about millions
same rate of return. The government would guar- of Medicare and Medicaid participants’ health
antee that, at retirement, the account balance conditions because it’s paying their medical bills.
would equal at least what the worker had con- This information has never, to my knowledge,
tributed, adjusted for inflation; that is, the govern- been inappropriately disclosed.
ment would guarantee that workers could not The vouchers would pay for basic in- and out-
lose what they contributed. This would protect patient medical care, prescription medications,
workers from the inevitable downside risks of and long-term care over the course of the year. If
investing in capital markets. you ended up costing the insurance company
Between ages 57 and 67, account balances more than the amount of your voucher, the insur-
would be gradually sold off each day by the Social ance company would make up the difference. If
Security Administration and exchanged for you ended up costing the company less than the
inflation-protected annuities that would begin voucher, the company would pocket the differ-
paying out at age 62. By age 67, workers’ account ence. Insurers would be free to market additional
balances would be fully annuitized. Workers who services at additional costs. The MSS would, at
died prior to age 67 would bequeath their account long last, promote healthy competition in the
balances to their spouses/partners or children. insurance market, which would go a long way to
Consequently, low-income households, whose restraining healthcare costs.
members die at younger ages than those of high- The beauty of this plan is that all Americans
income households, would be better protected. would receive healthcare coverage and that the
Finally, under this reform, neither Wall Street nor government could limit its total voucher expen-
the insurance industry would get their hands on diture to what the nation could afford. Unlike
workers’ money. There would be no loads, no the current fee-for-service system, under which the
commissions, and no fees. government has no control of the bills it receives,

F E D E R A L R E S E R V E B A N K O F S T . LO U I S R E V I E W J U LY / A U G U S T 2006 247
Kotlikoff

the MSS would explicitly limit the government’s oldest boomers will be eligible for early Social
liability. Security benefits. In six years, the boomer van-
The plan is also progressive. The poor, who guard will start collecting Medicare. Our nation
are more prone to illness than the rich, would has done nothing to prepare for this onslaught of
receive higher vouchers, on average, than the rich. obligation. Instead, it has continued to focus on
And, because we would be eliminating the current a completely meaningless fiscal metric—“the”
income-tax system, all the tax breaks going to the federal deficit—censored and studiously ignored
rich in the form of non-taxed health-insurance long-term fiscal analyses that are scientifically
premium payments would vanish. Added together, coherent, and dramatically expanded the benefit
the elimination of this roughly $150 billion of levels being explicitly or implicitly promised to
tax expenditures, the reduction in the costs of the baby boomers.
hospital emergency rooms (which are currently Countries can and do go bankrupt. The United
subsidized out of the federal budget), and the States, with its $65.9 trillion fiscal gap, seems
abolition of the huge subsidies to insurers in the clearly headed down that path. The country needs
recent Medicare drug bill would provide a large to stop shooting itself in the foot. It needs to adopt
part of the additional funding needed for the MSS generational accounting as its standard method
to cover the entire population. of budgeting and fiscal analysis, and it needs to
adopt fundamental tax, Social Security, and
Eliminating the Fiscal Gap healthcare reforms that will redeem our children’s
A 33 percent federal retail-sales tax rate would future.
generate federal revenue equal to 21 percent of
GDP—the same figure that prevailed in 2000.
Currently, federal revenues equal 16 percent of
REFERENCES
GDP. So we are talking here about a major tax hike. Auerbach, Alan and Oreopoulos, Philip. “The Fiscal
But we’re also talking about some major spending Effects of U.S. Immigration: A Generational
cuts. First, Social Security would be paying only Accounting Perspective,” in James Poterba, ed.,
its accrued benefits over time, which is trillions Tax Policy and the Economy. Volume 14. Cambridge,
of dollars less than its projected benefits, when MA: MIT Press, 2000, pp. 23-56.
measured in present value. Second, we would be
Fehr, Hans; Jokisch, Sabine and Kotlikoff, Laurence J.
putting a lid on the growth of healthcare expen-
“The Role of Immigration in Dealing with the
ditures. Limiting excessive growth in these expen-
Developed World’s Demographic Dilemma.”
ditures will, over time, make up for the initial
FinanzArchiv, September 2004, 60(3), pp. 296-324.
increase in federal healthcare spending arising
from the move to universal coverage. Third, we’d Fehr, Hans; Jokisch, Sabine and Kotlikoff, Laurence J.
reduce federal discretionary spending by one-fifth “Will China Eat Our Lunch or Take Us to Dinner?
and, thereby, return to the 2000 ratio of this spend- Simulating the Transition Paths of the U.S., the
ing to GDP. Taken together, these very significant EU, Japan, and China.” NBER Working Paper No.
tax hikes and spending cuts would, I believe, 11668, National Bureau of Economic Research,
eliminate most if not all of our nation’s fiscal gap. October 2005.

Ferguson, Niall and Kotlikoff, Laurence J. “Benefits

CONCLUSION Without Bankruptcy—The New New Deal.” The
New Republic, August 15, 2005.
There are 77 million baby boomers now rang-
ing from age 41 to age 59. All are hoping to collect Gokhale, Jagadeesh and Smetters, Kent. “Measuring
tens of thousands of dollars in pension and health- Social Security’s Financial Problems.” NBER
care benefits from the next generation. These Working Paper No. 11060, National Bureau of
claimants aren’t going away. In three years, the Economic Research, January 2005.

248 J U LY / A U G U S T 2006 F E D E R A L R E S E R V E B A N K O F S T . LO U I S R E V I E W
 Kotlikoff

Gokhale, Jagadeesh and Smetters, Kent. Fiscal and

Generational Imbalances: New Budget Measures
for New Budget Priorities. Washington, DC: The
American Enterprise Press, 2003.

Gokhale, Jagadeesh; Kotlikoff, Laurence J. and

Sluchynsky, Alexi. “Does It Pay to Work?”
Unpublished manuscript, January 2003.

Hagist, Christian and Kotlikoff, Laurence J. “Who’s

Going Broke?: Comparing Healthcare Costs in
Ten OECD Countries.” Milken Institute Review
(forthcoming).

Heckman, James; Lochner, Lance and Taber,

Christopher. “Explaining Rising Wage Inequality:
Explanations with a Dynamic General Equilibrium
Model of Labor Earnings with Heterogeneous
Agents.” Review of Economic Dynamics, January,
1998, 1(1), pp. 1-58.

Kotlikoff, Laurence J. Generational Policy. Cambridge,

MA: MIT Press, 2003.

Siegel, Jeremy J. The Future for Investors: Why the

Tried and the True Triumph Over the Bold and the
New. New York, NY: Crown Publishing Group,
2005.

F E D E R A L R E S E R V E B A N K O F S T . LO U I S R E V I E W J U LY / A U G U S T 2006 249
250 J U LY / A U G U S T 2006 F E D E R A L R E S E R V E B A N K O F S T . LO U I S R E V I E W