February 3, 2003
By William G. Shipman
logic supporting the reform of Social Security is to avoid the
converging paths of two global forces, which are proceeding on a
collision course. The first is the aging of society. This does not
mean that each of us is getting older, which is true, but rather
that the elderly population is increasing more rapidly than the
population as a whole. The second is that Social Security systems,
which provide most of elderly people's financial support, are not
sustainable as they are presently structured.
paths that these forces ultimately take will affect most countries
of the world--both developed and lesser developed--the prosperity
of workers, the security of the elderly, and people's very economic
well-being. Much is at stake, and the challenges are real--but the
opportunities are unprecedented.
few as five years ago, discussing the implications of population
aging on the financial soundness of state-run Social Security
systems was not common in public discourse. Indeed, addressing the
issue was often considered to be in bad taste, in some venues even
the last half decade, this has all changed. The financial integrity
of most Social Security systems is now subject to frequent
commentary across much of the world. The debate, although with
differences across countries, is whether Social Security should be
fundamentally reformed or whether the status quo, with minor
changes, should prevail.
Social Security is a pay-as-you-go system.
Unlike a privately funded plan where contributions are invested in
wealth-producing assets for retirement, Social Security taxes are
immediately paid out as benefits. Should taxes exceed benefits,
which has been the case in the U.S. since 1983, the surplus is not
explicitly saved or invested for future payments.
Today's benefits to the old are paid by
today's taxes from the young. Tomorrow's benefits to today's young
are to be paid by tomorrow's taxes from tomorrow's young. Thus,
benefit payments are an intergenerational transfer of wealth from
younger workers to older retirees.
total amount that can be paid in benefits is the payroll subject to
tax times the tax rate applied to that payroll. Holding the tax
rate constant, benefits can increase year-to-year by no more than
payroll increases. In the United States, that increase over the
last four decades has been about 1.6 percent per annum, adjusted
for inflation and increases in the work force.
is the rough equivalent of saving and investing for retirement and
earning an annual real rate of return of 1.6 percent. Investing
$1,000 at the end of each year for a 44-year working career at this
average rate of return would accumulate to about $63,000.
the last three quarters of a century, a portfolio of stocks and
bonds--with a 60/40 percent allocation, respectively--earned an
annual average 5.5 percent real rate of return. At this rate, the
yearly $1,000 saving would have accumulated to about $172,000. A
portfolio of only stocks, earning 7.4 percent annually, would have
grown to about $296,000. Neither capital market returns nor the
accumulated wealth from investing in them were without risk, but
this simple comparison is at least a first glimpse that compares
pay-as-you-go to market-based financing.
The Role of
Keeping payroll taxes and benefits reasonable in a
pay-as-you-go structure requires many workers to tax for each
individual who receives benefits. This ratio is primarily
determined by life expectancy and the fertility rate: the average
number of births per woman during her childbearing years. If life
expectancy increases or the fertility rate falls, the relative
number of workers declines.
expectancy is influenced in part by wealth. People who live in
wealthy countries generally experience longer lives than people who
live in poor countries. Thinking about it, it makes sense. Wealthy
countries can more easily afford clean water, clean air, better
medical facilities and access to them, the benefits of advanced
pharmacology, and the like: the things that make it easier for all
of us to survive. In the World Bank's path-breaking book, Averting
the Old Age Crisis, the authors state:
The proportion of the population that is
old rises with per capita income. In low-income countries, less
than 7 percent of the population is over 60. This proportion rises
to 12 to 16 percent in middle-income countries to 17 percent or
more in most high-income countries. The ratio of old people to
working age people (the old age dependency ratio) also rises with
per capita income--a relationship that stems directly from the
lower fertility rate in richer countries and the ability to
lengthen life span through medical intervention.1
fertility rate that ultimately stabilizes the size of a population
is called ZPG or zero population growth. It is 2.1. That is, if
women had 2.1 children, the population eventually would neither
grow nor shrink. As stated above, poor countries have higher
fertility rates than wealthy countries.
sharp contrast, consider data from the Central Intelligence
Agency's The World Factbook 2001.2 Wealthy (GDP per capita)
Switzerland ($28,600) has a fertility rate of only 1.47, whereas
poor Mozambique ($1,000) has a fertility rate of 4.82. Most of the
world's countries lie roughly between Switzerland and Mozambique in
both metrics of wealth and fertility rates.
Interestingly, however, abnormally low
fertility rates prevail throughout much of wealthy Europe and
Japan. For example, note: Germany, 1.38; France, 1.75; Italy, 1.18;
Spain, 1.15; Japan, 1.41, while the range of GDP per capita for
these countries is $18,000 to $24,900.3 As has been pointed out by
others, "there is now no single country in Europe where people are
having enough children to replace themselves when they die." While
these low fertility rates challenge the very survival of Europe's
Social Security systems, public pensions account for as much as 80
percent of total retiree income.4 Over time and without
change, Europe's elderly will face increasing insecurity.
interesting paradox is that as countries become more wealthy, their
pay-as-you-go pension systems become more frail because of the
relationship between wealth, fertility rates, and life expectancy.
As time passes, low fertility rates and increased life expectancy
cause the number of working-age people to fall relative to those
over 65 years of age. According to a study by the Center for
Strategic and International Studies, there will be a marked decline
in this ratio for many of the world's richest countries.5
Forecasts of the Number of Working-Age
for Each Person Over Age 65
in mind, this is not the same statistic as the number of workers
for each person receiving Social Security benefits, but rather the
number of working-age people. The number of workers actually paying
taxes would in almost all cases be fewer.
These forecasted trends do not counter
history. Over the last half-century, the ratio of workers to
beneficiaries already has declined. For example, in the U.S. there
were 16 workers per beneficiary in 1950; today there are only 3.4.
The above forecasts are just a continuation of this precedent.
Because of past demographic realities and in order to pay
promised benefits, governments have most often responded by raising
taxes, slowly--parallel to the chelonian change in aging--but
relentlessly. For example, in the United States, the maximum
payroll tax levied just for the retirement portion of Social
Security, so-called Old Age and Survivors Insurance, increased from
$90 to $8,999.40 over the last five decades, an inflation-adjusted
jump of about 1,200 percent.6 And now Social Security's
actuaries report that there will be as few as 2.1 workers per
beneficiary by 2030.7
history is our only guide, payroll taxes will increase further.
Already, workers in European countries pay significantly higher
Social Security taxes than their American counterparts even though
for about 75 percent of all American workers Social Security taxes
are greater than their income taxes.
At some point, the strategy of raising taxes hits a
political wall. People sense that their payroll taxes will not
provide them the security that they could acquire elsewhere with
the same level of resources. The tax burden of concern differs
across countries and ideologies. Some people are less indifferent
to taxes than others. Many countries have long histories of high
tax burdens that would not be tolerated in other countries.
These important differences aside, we may
now be facing this so-called political wall almost everywhere. One
support of this view is the fact that the idea of fundamentally
reforming Social Security is now common discourse across the globe
including in the United States, Russia, China, Japan, and the
European Union. New solutions are being sought.
of the often-proffered suggestions is reducing benefits. Such
reductions are usually expressed in rather opaque terms such as
raising the eligible age to receive benefits, increasing the years
of wage income necessary to receive full benefits, establishing
initial benefits to price instead of wage indexing, adding or
changing bend points in benefit formulae, pegging benefits to life
expectancy instead of a specific retirement age, and so on.
Reducing benefits, if applied in
isolation, is the flip side of raising taxes. It approaches the
financial challenge in strictly cash flow terms. From this narrow
perspective, raising taxes or reducing benefits makes sense. But
neither appropriately addresses broader long-run societal,
financial, or economic concerns, and neither changes
below-replacement birth rates or the increasing of life expectancy.
Reducing benefits, much like raising taxes, is a financial
Another solution, which has only recently
gained ground, is the fundamental restructuring of pay-as-you-go
systems to market-based financing. That is, paying less in Social
Security taxes and investing that amount in wealth-producing assets
such as stocks and bonds.
Strictly moving to market-based financing,
however, is not a panacea. Market-based systems can be poorly
designed, less efficient than pay-as-you-go systems, and riddled
with risk, and can ultimately fail. It is important that their
design take into account all the benefits of markets while not
incurring the costs of a poor administrative structure.
Countries that eventually employ market-based financing
probably will adopt some common design characteristics, although no
two countries will likely settle on exactly the same administrative
structure. One reason is national pride; another is that their
starting points differ. For instance, the development of capital
markets is well-established in some countries and nascent in
others. This effects the decision on whether portfolios may invest
across borders or not. This in turn influences investment risk.
Computer technology, availability of
hardware, and software applications are not uniform across
countries. This will influence the timing of investment flows,
portfolio pricing, the ability to change asset classes in an
individual's portfolio, and whether individual portfolios are even
Personal property rights are
well-established in some countries and not in others. In the latter
case, this may require that the government owns the assets and,
therefore, makes all investment and benefit decisions. From an
administrative point of view, this is a more simple design than
that required for individual accounts.
In most countries, the above administrative issues have
not been fully explored. The United States, however, is an
exception. During the past seven or so years that Social Security
reform has been debated here, an administrative platform has been
designed. It was presented in testimony before the House Budget
Committee Social Security Task Force in April 1999.8 Subsequent
to the testimony, the committee's chairman, Congressman Nick Smith
(R-MI), wrote a letter to David Walker, Comptroller General of the
United States, in which he stated:
The House Budget Committee Task Force on
Social Security recently heard testimony from William Shipman....
His firm completed a study that concluded that the administrative
annual costs for establishing broadly diversified Social Security
personal retirement accounts would equal between $3.38 and $6.58
per account holder. If verified, I believe the conclusion will
prove highly significant as Congress evaluates plans to modernize
the retirement system. It will demonstrate that it is possible to
provide meaningful investment opportunities to all Americans for
only 1 or 2 cents a day. For that reason, I am requesting that the
[U.S. General Accounting Office] study the methods and conclusion
of this report to determine its accuracy.
GAO report9 was published during June
1999. The administrative plan was also reviewed by the President's
Commission to Strengthen Social Security and in large part adopted
by the commission in its December 11, 2001, report.
design, arguably the most thorough to date, was conceived by a
working group of academics, investment and recordkeeping
professionals, actuaries, and individuals involved in both policy
and politics. Almost all in the group had been involved in Social
Security issues for years, in some cases as long as 20 years. This
was particularly helpful because the group was able to consider not
only narrow technical points, but broader political and policy
considerations as well.
goal was to design a platform for an individual-account,
market-based Social Security option. It considered eight necessary
These objectives were considered important
because they had been central in the U.S. debate on Social Security
reform. They are also integral to the most popular defined
contribution system in the United States, the 401(k) plan.10 Indeed,
the 401(k) plan structure is often referenced as a potential model
for an individual retirement account Social Security plan.
The 401(k) Plan:
History and Structure
Since the late 1970s, defined contribution systems have
increased in popularity among American companies and workers. And
just since 1985, those that have sponsored as well as those that
have participated in 401(k) plans have increased several fold.
Under 401(k) programs, a plan sponsor,
usually a company or union, oversees administration of a savings
and investment program for its employees. Under such plans:
the early years of 401(k) plans, investment options were often
limited to a stable value fund, a diversified fund, and company
stock. In recent years, however, there has been a significant
increase in investment choice. Many plans now include a large
number of investment options as well as self-directed brokerage
accounts. These accounts provide access to a large universe of
institutional and mutual funds as well as individual securities.
With all of the choice available, individuals can now create
portfolios that are appropriate for their age, their risk
tolerance, and their wealth objectives.
Security/401(k) Plan" Challenge
The major challenge in creating a 401(k) model of
individual accounts linked to Social Security is the timely posting
of individuals' savings contributions. This is not possible given
the present Social Security recordkeeping system.11 Although
the Treasury Department has built a comprehensive system for the
collection of FICA taxes from employers, there is no detailed
record of individual taxes paid at the time they are collected and
sent to Treasury. This information is not communicated to the
government until the following year.
Companies remit FICA taxes in lump sums
throughout the calendar year but do not forward to the government
at the same time the names of the individual employees who paid
those taxes or the amount each paid. That information is not
provided to the government until the next calendar year when the
employer sends W-2 forms to both the government and the
Treasury knows throughout the current
year, for instance, that a company has paid a sum of FICA taxes for
its employees, but the Social Security Administration will not
update its records until June or so of the following year, and
possibly a few months later, with the names of the individual
workers who paid those taxes and how much each worker paid. The
government never knows when during the year the individual paid the
recordkeeping process, although workable in Social Security's
defined benefit structure, is unworkable in a daily-environment
defined contribution structure. But it is all that currently exists
for identifying individual payroll taxes.
solution to this recordkeeping problem is to structure investment
options, not all of which require timely and detailed contribution
data. This approach involves three investment levels.
Investment: A Pooled Money Market Account
At the first level, workers' savings are deducted from
payroll and invested in a collective money market fund. Workers own
the assets of the fund, although the accounting at the individual
level is not completed until the following year. This
reconciliation is accomplished with the filing of the W-2 form.
When the individual's assets are accounted for, units in the money
market fund, which include earned interest, are then posted to each
worker's account. The fund is dollar priced, which means each unit
is always valued at one dollar.
units are then invested in one of three balanced funds selected by
the worker. Individuals who do not or choose not to make a
selection have their assets invested in a default option, which is
one of the balanced funds. The account holder has the option--after
a startup of about three years, a period required to successfully
build up assets to achieve economies of scale--to transfer some or
all of his balance to an appropriate retail retirement account.
pooled account would be a conservative fund similar to a large
institutional money market fund. The funds would be held in this
pool earning interest for all participants.
Given that the timing of an individual's
contribution is not known, all participants are assumed to invest
on June 30. The effect of this recordkeeping accommodation is that
interest credited to one's account could be more or less than what
was actually earned. In most cases, the differences are trivial.
For an average-income individual who works continuously throughout
the year, there is no effect. High-income workers, however, would
effectively subsidize other workers because high-income workers
would contribute a disproportionate amount of their income during
the early part of the year. Alternatively, individuals who enter
the labor force after mid-year would benefit from this
worker would know during the year how much is invested because it
is the same as the year-to-date reduction in the FICA tax that goes
to savings, often referred to as the carve-out. The carve-out may
be itemized as a separate line item on the pay stub. Interest would
always accrue, so the account balance would be in excess of the
contribution. All workers, regardless of income, would receive an
identical rate of return. Funds would remain in the money market
account until the reconciliation date of how much each worker
contributed--about August of the following year.
Investment: Balanced Funds
When the individual account balance is determined, it is
invested in one of three balanced funds that the worker chooses.
Balanced funds are diversified portfolios that are generally
invested in stocks, bonds, and cash.
combined assets underlying very successful private,
employer-sponsored defined benefit plans are essentially balanced
funds. One of the Level Two balanced funds may have an allocation
that closely approximates these plans. This allows all workers, if
they wish, to maintain an asset allocation similar to that provided
to the employees of many sophisticated corporations in the world.
This would be the default fund. There would be another fund on each
side of this fund: one for younger workers that would be weighted
more toward equities, while the other would be weighted more toward
bonds for those closer to retirement. Irrespective of their age,
however, workers could choose any of the three funds.
Although workers could choose their
balanced fund, some may not. In this case, they would default to
the middle fund. In other words, a worker--perhaps low-income and
financially unsophisticated--would be invested in a highly
diversified balanced portfolio suited for retirement savings. The
portfolios would be managed by professional asset managers chosen
through an open and competitive bidding process.
Index fund investment management fees most
likely would be less than two basis points: two one-hundredths of
one percentage point. The balanced funds would be valued daily, and
prices would be published in the popular press. Workers only need
multiply their units, which are a constant for one year, by the
daily price to monitor their account balance.
Investment: Rollover Option
After a period of perhaps three years--a period required
to successfully build up the assets in the Level Two account system
to realize economies of scale--investors seeking more investment
choice would have the option of rolling their investment funds out
of the Level Two asset allocation funds and into any qualified
retirement investment account.
Those choosing Level Three would transfer
assets to a qualified account with a financial services company
meeting reasonable and specified standards.12 While investors would have
a wider range of choice within Level Three, there still would be
reasonable investment guidelines. In Level Three, investment
managers would act as the fiduciary for their product offerings and
be subject to Department of Labor oversight. This is consistent
with many employer-sponsored plans, both defined contribution and
Level Three might well suit those workers
who have a high degree of confidence in a particular money manager,
a particular firm, or a particular style of investing. It will also
serve investors seeking a type of investment unavailable in the
Level Two asset allocation funds. An investor, for example, may
wish a greater concentration of equity investments than is
available in the asset allocation funds.
Should a worker find a particular Level
Three provider or product unsatisfactory, the worker could transfer
to another provider within Level Three or move back to Level Two.
This assures competition across the Level Three retail platform as
well as competition between Level Two, an institutional platform,
and Level Three. This competition within and between Levels will
ensure the lowest cost and best service for the entire system.
The administration of an individual account system will
require the development of a large-scale, customized recordkeeping
system with the capability to produce a highly efficient service
solution. The efficiency of the service application will be
dependent upon the design and execution of the system. There is no
existing application that meets all the requirements.
requirements to support a national individual account system will
be complex, large-scale, and capital-intensive. As noted above,
this is a challenge of unprecedented scope. Nonetheless, the
application itself is relatively straightforward.
Development time can be minimized to allow
focus on sizing and scaling the network and building the necessary
interfaces to the Social Security Administration. Unlike mutual
fund or 401(k) recordkeeping systems, there will not be many unique
product features or functions, thus significantly reducing
complexity and cost. It is reasonable to assume a system could be
developed in 12-18 months to support these requirements.
greatest challenge in building a recordkeeping system to support
the requirements of an individual account system is not the
complexity of the application, but the need to support the high
volume of participant inquiries, transactions, transfers, and
report generation. To keep costs low, it is critical that most
participants utilize voice and Internet technology to obtain
information and transact business. The greater the percentage of
calls that requires a customer service agent, the higher the
volume of calls will be driven by the frequency of transactions and
statements, as well as average account size and market volatility.
Assuming 140 million accounts and the plan described, participant
call volumes are projected to range from 175 million to 350 million
annually. In addition, the system will issue 140 million
statements, process fund transfers, and distributions. This
approach assumes the funds are priced daily and accounts updated
Whether the recordkeeping is done by a
government entity such as the Social Security Administration or
outsourced to the private sector, this task will require the
formation of a large service organization to support these
requirements. The service firm would need call centers in multiple
locations around the country and would need to hire between 3,000
and 7,000 employees. For the purpose of this analysis, it is
assumed that the Social Security individual account system will
incur volumes between 0.5 and 1.0 calls per participant per
Another important factor in projecting
costs is determining what percentage of the participant's call
volume will be processed by voice response and Internet technology
versus requiring the services of a call center representative. The
cost to handle calls using the technology is a fraction of the cost
to process through a representative. Therefore, to achieve an
efficient solution, it is critical that high levels of automation
are achieved. The analysis assumes 85 percent of the call volume
will be handled through voice and Internet technology, comparable
to the levels currently achieved in many large defined contribution
Based on the plan design defined above, a cost model has
been developed to project the administration costs under a range of
assumptions. The unit cost factors are based on experience in the
401(k) business and have been adjusted in some cases to account for
the scale of the individual account option.
requirements of a national system of individual accounts are
unique, and, therefore, extrapolations from 401(k) experience pose
some risks. Unlike the 401(k) structure, it is assumed that in a
timely fashion the Social Security Administration will provide the
individual account recordkeeper an accurate, automated transmission
of earnings histories that will be used to calculate annual
These and any other expenses associated
with reconciling W-2 records are to be borne by Social Security and
are not included in this cost model. It is also assumed that Social
Security, at its cost, will maintain accurate and up-to-date
employee address files, as will be necessary anyway with the annual
mailing of Social Security statements starting in 2000. One's
investment account statement would be included in this mailing.
Another cost not included in this model is
the expense associated with communicating this program to
employees. The assumption is that the government would bear these
expenses. Therefore, they expressly are not included in the
asset-based fees reported below. There is precedent for this in
that the government pays directly some of the communication expense
of the Federal Thrift Plan.
Based on the design criteria outlined and unit cost
assumptions, it is projected that the first year's administrative
expenses to support an individual account system will range from
$473 million to $922 million. Assuming 140 million accounts, this
translates to a cost per account range of $3.38 to $6.58 in the
Although costs would be expected to
increase annually, driven primarily by employee compensation and
benefits, assets would increase more rapidly. Costs as a percent of
assets, therefore, would fall. Steady-state asset-based costs are
projected to range from 20 to 40 basis points.
These costs are competitive with other
investment products. For example, the Federal Thrift Plan, which is
often used as an example of an efficient retirement plan, had an
expense ratio of 65 basis points in its second year. Another
benchmark is a portfolio of individual mutual funds representing
different asset classes and weighted to approximate a Level Two
balanced fund. Such a portfolio is presently available for a total
cost of about 40 basis points.14
Although many approaches to the
administrative challenges inherent in an individual account system
linked to Social Security may be expensive, not all need to be.
Under reasonable assumptions, a well-thought-out plan that meets
our nation's retirement needs is affordable.
debate surrounding the future of Social Security systems around the
world has attracted a wide range of views and opinions. Although in
most cases final designs are far from certain, and perhaps far off,
open dialogue is a necessary precursor to their enactment.
These discussions over many years have
focused policymakers and government leaders on the problem, the
significant challenge, the stark implications of doing nothing, and
the substantial benefits from thoughtful and successful reform. The
willingness of all interested parties to engage in this debate will
help more Americans and leaders of other nations understand the
importance of solving one of the world's most challenging
William G. Shipman is Chairman of CarriageOaks
Partners LLC, a Massachusetts-based consulting firm specializing in
retirement finance. He spoke at the Inaugural Symposium on
Financing Global Aging, sponsored by the Rosenberg Institute of
Global Finance, Brandeis University, Graduate School of
International Economics and Finance.
Show references in this report
1. World Bank, Averting
the Old Age Crisis (New York: Oxford University Press, Inc., 1994),
2. Data are from The World
Factbook 2001 at
4. Richard Jackson, The
Global Retirement Crisis: The Threat to World Stability and What to
Do About It (Washington, D.C.: Center for Strategic and
International Studies, 2002), p. 25.
5. Robert Stowe England,
The Fiscal Challenge of an Aging Industrial World (Washington,
D.C.: CSIS Press, 2002), p. 50. The data are transformed from
old-age dependency ratios.
6. Author's calculations
from The 2002 Annual Report of the Board of Trustees of the Federal
Old Age and Survivors Insurance and Disability Insurance Trust
Funds, Table VI.A1, pp. 126-127.
7. Ibid., Table IV.B2, p.
8. For a more complete
analysis, see "Testimony of William G. Shipman Before the House
Budget Committee Task Force on Social Security," April 27,
10. Profit Sharing/401(k)
Council of America, "Helping Americans to Help Themselves: The Role
of Profit-Sharing/401(k) Plans in the Retirement-Income Security
11. See, for example, Kelly
A. Olsen and Dallas L. Salisbury, "Individual Social Security
Accounts: Issues in Assessing Administrative Feasibility and
Costs," Employee Benefit Research Institute Special Report No.
SR-34 and Issue Brief No. 203, November 1998.
12. This process should be
fully automated and driven by a third party, such as the National
Securities Clearing Corporation. At the individual account holder's
instructions, the Level Three provider should be able to initiate
the transfer and cause the money to be moved from Level Two to
Level Three without having to provide any paperwork.
13. U.S. Department of
Labor, Pension and Welfare Benefits Administration, "Study of
401(k) Plan Fees and Expenses," April 13, 1998.
14. Melissa Hieger and
William Shipman, "Common Objections to a Market-Based Social
Security System: A Response," Cato Institute Social Security Paper
No. 10, July 22, 1997, p. 13.
One solution to the significant problem of recordkeeping, forexample, would be to structure investment options, not all of whichrequire timely and detailed contribution data.
William G. Shipman
Read More >>
Heritage's daily Morning Bell e-mail keeps you updated on the ongoing policy battles in Washington and around the country.
The subscription is free and delivers you the latest conservative policy perspectives on the news each weekday--straight from Heritage experts.
The Morning Bell is your daily wake-up call offering a fresh, conservative analysis of the news.
More than 200,000 Americans rely on Heritage's Morning Bell to stay up to date on the policy battles that affect them.
Rush Limbaugh says "The Heritage Foundation's Morning Bell is just terrific!"
Rep. Peter Roskam (R-IL) says it's "a great way to start the day for any conservative who wants to get America back on track."
Sign up to start your free subscription today!
The Heritage Foundation is the nation’s most broadly supported public policy research institute, with hundreds of thousands of individual, foundation and corporate donors. Heritage, founded in February 1973, has a staff of 275 and an annual expense budget of $82.4 million.
Our mission is to formulate and promote conservative public policies based on the principles of free enterprise, limited government, individual freedom, traditional American values, and a strong national defense. Read More
© 2014, The Heritage Foundation Conservative policy research since 1973