The Big Lie About the “Entitlement State“

In 1960, government transfers to individuals totaled $24 billion. By 2010, that total was 100 times as large. Even after adjusting for inflation, entitlement transfers to individuals have grown by more than 700 percent over the last 50 years. … There are sensible conclusions to be drawn from these facts. You could say that the entitlement state is growing at an unsustainable rate and will bankrupt the country.

—David Brooks, “Thurston Howell Romney,” New York Times, September 17, 2012

Is the view that “entitlements”—government programs like Social Security, Medicare, and Medicaid—“will bankrupt the country” a “sensible conclusion”? No. It’s scare-mongering of the “OH MYGOD WE’RE ALL GOING TO DIE!” variety, completely unjustified by a sober look at data on government transfer payments between 1960 and 2010.

New York Times columnist David Brooks starts the passage on entitlements in his September 17 column by noting that total government transfer payments have increased by an alarming-sounding 100 times over the last half-century. In the next sentence, he acknowledges that this figure is not adjusted for inflation. (Nor for population growth.) As it turns out, the “100 times” mostly reflects the increase in the general price level (more than seven-fold between 1960 and 2010) and the growth of the U.S. resident population (not quite doubled), not the growth in transfer programs specifically. Correcting for these factors, Brooks admits, the increase is just “700 percent.” One can only guess that he switched to percentage terms because he’s trying to sound scary, and “700 percent” sounds far scarier than “seven times.” (Brooks actually describes this figure simply as “after adjusting for inflation,” but it appears that he actually adjusted for both inflation and population growth.)

That’s as far as Brooks gets, so he misses another crucial adjustment. The average income in the United States is far greater today than it was in 1960. Real GDP per capita grew by more than two-and-a-half times between 1960 and 2010. Now, looking at real entitlements spending per capita relative to real GDP per capita (or just real entitlements spending relative to real GDP), the growth over the last 50 years is down to less than three-fold. It makes perfect sense that cash benefits programs like Social Security—which send people checks and allow them to spend the money as they see fit—should grow with increasing incomes. These programs are meant to help people maintain something resembling the customary standards of living of today, after all, not those of the Eisenhower era.

With a few sensible adjustments, then, Brooks’ alarming initial figure of “100 times” vanishes almost into thin air. That still leaves, however, an increase of a little less than three times. What accounts for that?

To begin with, over 70% of the increase in social benefits at all levels of government, over the half century between 1960 and 2010, is accounted for by three programs: Social Security, Medicare, and Medicaid. Two of these, Medicare and Medicaid, did not even exist in 1960. (Social Security, meanwhile, did not cover anywhere near the percentage of the labor force it covers today.) It is rather disingenuous to bemoan the “unsustainable growth” of certain government programs, over a certain period, when they did not even exist at the beginning of that period.

More generally, the growth in the Big Three social-benefits programs is the combined result of several different effects. Partly, it reflects changes in the demographic composition of the population. Social Security and Medicare primarily benefit the elder population. This age group has grown as a percentage of the overall U.S. population because people are, on average, living longer and because the demographic “bump” of the baby-boom generation is beginning to reach retirement age. Partly, the increase reflects the growth in medical costs, which has been faster than the increase in the general price level. Finally, it reflects the expansion of benefits associated with these programs (Social Security retirement benefits, for example, are tied to lifetime earnings, so as earnings go up so do benefits).

Even aside from the numbers, Brooks is fundamentally wrong that transfer programs can “bankrupt the nation.” Transfer programs, as their name suggests, transfer income from one part of the population to another. Social Security, for example, is primarily an intergenerational transfer program. It taxes current workers to fund benefits for current retirees. (Most people pay taxes that fund other people’s benefits, during one part of their lives, and then receive benefits paid for by other people’s taxes, during another part.) The “losses” for those who are paying the taxes, at any given time, are not losses to society as a whole. They are balanced by the gains to those who are receiving the benefits.

In another sense, the benefits to everyone covered by these programs greatly exceed just the cash amount of the transfers. Social Security provides not only a retirement annuity (insurance against destitution during old age), but also disability benefits (insurance against being unable to work) and survivors’ benefits (insurance for family members against the death of a financial provider). Medicare and Medicaid, meanwhile, pay for medical services and prescription medicines. These programs, in short, offer meaningful protection against many of life’s possible calamities.

When we think of costs (of a government program or of something else) to “society as a whole,” we need to think about the use of real resources—a part of society’s total labor time, buildings, tools, etc.—that could have been used for some other purpose. The real costs of the Big Three transfer programs, in this sense, fall into two categories.

First, there are costs involved in administering the programs. The work hours and other resources (office buildings, desks, chairs, computers, electricity, pencils, paper clips, etc.) used to keep program records, send out benefits checks, and so on, could have been used for something else. So those represent real costs to society. In the case of the major transfer programs, however, the administrative costs are very small relative to the total benefits paid. The costs of administering Social Security, for example, are less than one percent.

Second, there are real costs for the goods and services for which government transfers pay. Medicare and Medicaid, for example, pay medical practices, hospitals, medical-supply companies, and pharmaceutical manufacturers to deliver medical care and medicines to program beneficiaries. Total transfer program costs, therefore, have increased along with rising medical costs. Part of the reason for rising medical costs has been that people now receive medical services they once could not. Magnetic resonance imaging (MRI) scans, for example, were not widely available two decades age. Now they are. Another is the rising real incomes of (some) medical professionals and the burgeoning profits of pharmaceutical manufacturers. Perhaps the most important, however, is that the U.S. health-care delivery system has enormous administrative costs, far above those of other high-income countries.

Advocates of single-payer (public) health insurance point out that such a system could 1) rein in pharmaceutical costs by using the government’s purchasing power to negotiate lower prices and 2) dramatically reduce administrative costs by eliminating the crazy quilt of different private-insurance billing systems. (See Gerald Friedman, Funding a National Single-Payer System, March/April 2012; Gerald Friedman, “Universal Health Care: Can We Afford Anything Less?” July/August 2011). Maybe David Brooks should start a clamor about that.

Two obvious ways to pay for growing transfer programs, from a public-finance standpoint, are: First, keep government revenue the same, but change its uses. For example, the United States could fight fewer wars, have a smaller military, and buy less military hardware. Or it could liberalize laws on recreational drug use, and reduce spending on police, courts, and prisons. It could use some of the savings to fund the Big Three and other social programs. Second, increase government revenue. Contrary to current mythology, the U.S. population is not being taxed to the very limits of its endurance. Of thirty high-income OECD countries, the United States ranks dead last in total tax revenue (for all levels of government) as a percentage of GDP, at less than 25%. The figure is 30% or more for 24 of the 30 countries, and over 40% for eight.

Brooks acts as if budget issues are one-sided: a matter only of how much a particular program or combination of programs costs. This one-sided view is especially evident in U.S. political discourse on deficits, which politicians and commentators often frame as a problem of excessive spending. A budget deficit, however, is the difference between expenditures and revenue—it is an inherently two-sided issue—so looking at the expenditures side alone doesn’t help us understand the causes of deficits or the possible policy responses.

Could the U.S. government just raise more revenue, as a percentage of GDP, to pay for transfers that have grown as a share of GDP? Well, somehow a couple of dozen other countries seem to manage. So probably yes.

SOURCES: Bureau of Economic Analysis, National Income and Product Accounts, Table 3.1.Government Current Receipts and Expenditures (bea.gov); Bureau of Labor Statistics, Consumer Price Index—All Urban Consumers (Series ID: CUSR0000SA0) (bls.gov); Bureau of Economic Analysis, National Income and Product Accounts, Table 1.1.3. Real Gross Domestic Product, Quantity Indexes (bea.gov); U.S. Census Bureau, Population Estimates (census.gov); Social Security Administration, Actuarial Publications, Administrative Expenditures (ssa.gov); OECD Tax Database, Table A. Total tax revenue as percentage of GDP (oecd.org/tax).