Taxes, Social Security & the Politics of Reform
From the November 29, 2004 issue: A Reaganite plan for Reagan's heirs.
Nov 29, 2004, Vol. 10, No. 11 • By JOHN D. MUELLER
BY PUTTING INCOME-TAX and Social Security reform atop his second-term domestic agenda, President George W. Bush courageously zeroed in on the two most important fiscal issues facing American families for at least a generation. The basic choice is whether to retain the formula Ronald Reagan advocated and pursued in office--combining a broad-based, low-rate income tax with a pay-as-you-go Social Security system--or shift to taxing the narrower base of "consumption" at higher rates while replacing Social Security pensions with individual retirement accounts.
In a seminal February 1977 speech, Reagan cautioned against those "on the left or right . . . who worship the god of political, social, and economic abstractions, ignoring the realities of everyday life." He called for a "new Republican party," which, by applying traditional conservative principles to economic, social, and national security policy, would serve as the "political mechanism through which the goals of the majority of Americans can be achieved." The most striking political fact about Republican presidents and presidential candidates since Reagan's departure is that, while faring steadily better on social and national security issues by following the Reagan approach, they have fared steadily worse on economic policy by moving away from it. For example, President Bush's 2004 reelection resulted from a whopping 19-point advantage in social and national security issues, which outweighed an 18-point deficit on economic policy issues. Had Bush merely broken even on economic policy, he would have enjoyed a historic landslide. This gives a sense of the magnitude of the political upside if Republicans can recapture the success of the Reagan fiscal program.
Beneath the legal barnacles left by decades of hand-in-pocket relations between lawmakers and lobbyists, U.S. fiscal policy retains a discernible economic, moral, and political logic, which owes much to Abraham Lincoln (who introduced the first income tax in 1862), to the "maternalist" (Allan Carlson's apt term) group of women who designed the pay-as-you-go Social Security retirement system in the 1930s, and to Ronald Reagan, who led bipartisan efforts to balance Social Security taxes and benefits (in 1983) and to reform the income tax (in 1986). This logic can be stated simply: The most just, efficient, and popular way to pay for common goods like national defense and general government is an income tax with the broadest possible base, lowest possible tax rates, and most equal treatment of labor income (salaries, wages, fringe benefits) and property income (interest, capital gains, dividends, rent, royalties, etc.). This last point is the one that has bedeviled the post-Reagan Republican party, as it has repeatedly been tempted in the name of capital formation, investment, and growth (all good things!) to tax property income less heavily than labor income, rather than treating the two equally. This is why Republicans, for all their reputation as tax-cutters, have had almost nothing to say about the larger bite of payroll taxes over the last two decades.
The same logic dictates that narrower benefits require narrower funding. For example, transfer payments to persons or families (including Social Security and unemployment benefits) should be paid by taxes on labor income, and subsidies for property-owners by taxes on property (like gasoline taxes and vehicle tolls to fund roads and bridges). In my experience, these principles are the key to an economic policy that is legislatively successful and politically popular. Ignore them and plausible schemes startle their proponents by erupting in political flames--as has happened with all proposals so far for a flat tax or Social Security privatization (most recently in 1996 and 1999, respectively).
TWO CENTRAL ECONOMIC REALITIES of everyday life are as true in 21st-century America as when Aristotle first pointed them out in 4th-century B.C. Athens: Most wealth is owned by families living in households, not the government, and this wealth is of two kinds, people and property--or as University of Chicago economist (and later Nobel laureate) Theodore W. Schultz dubbed them in 1960, "human and nonhuman capital." But there's a key difference. Since the ownership of property can be subdivided almost indefinitely and transferred in competitive markets, everyone receives the same rate of return (other things, like risk, being equal). But human capital is embodied in people, for whom (since the abolition of slavery) there is fortunately no longer a market, and its rate of return varies with the age of the person in whom it is embodied. For example, in 1999-2000 the real rate of return on college tuition at age 20 was about 16 percent, compared with the stock market's long-term average return of 6 to 7 percent. But by age 40 your money would be better invested in property, and at about age 50 the return on tuition turns negative. (The main reason is that four years of college roughly double average annual earnings, but as we get older, those earnings can be realized for fewer years.)
Such differences between people and property have always existed, but two important changes have occurred mostly within the last century: the separation of the household and business firm, and the unprecedented increase in human lifespan. In December 1861, Abraham Lincoln could accurately report that, "In most of the southern States, a majority of the whole people of all colors are neither slaves nor masters; while in the northern a large majority are neither hirers nor hired. Men with their families--wives, sons, and daughters--work for themselves, on their farms, in their houses, and in their shops, taking the whole product to themselves, and asking no favors of capital on the one hand, nor of hired laborers or slaves on the other." In other words, human and nonhuman wealth were still mostly owned and produced by family households.
Since then, the two kinds of production have become specialized. The modern business firm, historically speaking, is a recent offshoot of the household and specializes in producing and maintaining productive property, while the modern household specializes in "producing" and "maintaining" human persons. One or both adults in every household (if not employed by a government or nonprofit organization) are typically employed by a business, which combines the services of workers and of productive property (buildings, land, raw materials, machines, patents) to produce new goods and services for sale in the market. Though most such property is directly owned by businesses, families still own it indirectly, by owning the shares and bonds issued by those companies. When the firm's product is sold--for example, when a family purchases a new car or computer--the money is paid out entirely as income to the producers: labor compensation (wages, salaries, and fringe benefits) to workers and property compensation (interest, dividends, and retained earnings or capital gains) to property-owners. In the final analysis, there is nothing else to tax.
The other key change is the unprecedented increase in the human lifespan. From ancient Rome to 15th-century Europe, human life appears to have averaged just 24 years. This meant that most people experienced only two phases of life, dependent childhood and active parenthood. As Angus Maddison has shown, it also meant there was no economic growth, because a high birth rate merely offset a high mortality rate, and most people didn't live long enough to accumulate more property than they used up. Steady advances in medicine and public hygiene increased life expectancy at birth to about 38 years in the United States in 1850 and 47 years in 1900. Since then, life expectancy at birth has risen by more years than in the previous five centuries combined--to about 80 for women and 75 for men. As a result, most people now experience four distinct phases of life: dependent childhood, active parenthood, the "empty nest," and retirement. Longer life has caused a sharp increase in investment in both people and property, and thus faster economic growth, because the returns can now be realized for many more years. These basic facts account for the pattern of lifetime earnings shown in the chart on the following page.
Early in life, income is mostly labor compensation, which starts at zero and increases as we acquire valuable skills; rises rapidly between childhood and the mid-30s as we enter and gain experience in the labor market; rises more slowly to peak at around age 50; then drops finally to zero in retirement. Property income starts close to zero early in life (for those with little or no inherited property), but becomes increasingly significant as the expected rate of return on investment in human capital falls below the rate on investment in property. And for those who acquire significant wealth from any source--whether inheritance, talent, luck, or hard work--the only practical way to save it is in the form of claims on property (stocks, bonds, etc.).
These facts of everyday life account for the distribution of income among American families. But to interpret them, we must reconsider the meaning of income. In the same article in which Schultz coined the term "human capital," his first policy conclusion was this: "Our tax laws everywhere discriminate against human capital. Although the stock of such capital has become large and even though it is obvious that human capital, like other forms of reproducible capital, depreciates, becomes obsolete, and entails maintenance, our tax laws are all but blind on these matters." What he meant was that labor and property income--the returns on investment in human and nonhuman capital, respectively--are measured differently. Before property income is ever taxed, the costs of maintaining the property in working order are excluded and a further allowance for the property's depreciation in use is deducted. Only what's left over after these calculations is taxed. But labor compensation is taxed without regard to its maintenance costs or depreciation. A farmer who buys a $50,000 tractor to increase the productivity of his operations will eventually deduct that full cost as well as maintenance and repairs from his income. Should he spend the same amount sending his daughter to ag school to become an expert manager of the family's property, he will enjoy no similar deductions.
It might be argued that the combination of standard tax deductions, personal exemptions, earned income and child tax credits, etc., which aim to exempt a poverty-level income from taxation, amount to a rough equivalent of minimal "human maintenance" costs. But there is no allowance for the depreciation of human capital, which, since we all die, is always 100 percent. Therefore a much larger share of labor than property compensation is subject to tax. The political process has clumsily and inefficiently responded by imposing progressive tax rates and multiple layers of taxation on property income.
The simplest way to treat all income equally is to add back the allowance for property depreciation to property income and to tax this larger amount, while subtracting a poverty-level income ("human maintenance") from labor compensation before taxing this smaller total. We can call the resulting measure "broad family income." About 75 percent of gross income for the lowest four-fifths of families is labor income, but the share falls to about 40 percent for the top 1 percent of families.
This is the crux of the tax-reform debate. Because "consumption"-tax advocates inconsistently define investment to mean investment in property but not in people, taxing "consumption" essentially means taxing labor compensation. For instance, in the example above, the tractor expenses would be treated as investment, the ag school expenses as "consumption" (which is why throughout this article I put the word "consumption" in quotation marks). I sometimes call this worldview the economic Stork Theory, because it begins with the assumption that people and their skills arrive from out of the blue, as if delivered by a large stork. Dick Armey's and Steve Forbes's flat-tax plan of the 1990s would have achieved this result by exempting from taxation both investment in productive property (through "expensing" of plant and equipment) and the personal income that is the return on that property (interest, dividends, rents, and capital gains). After a transition period, the result would have been to exclude all property income from the tax base--leaving, in effect, a glorified payroll tax. (A national sales tax gets the same result immediately.) Since property income is about one-third and labor income two-thirds of total national income, such a tax reform would require a new tax rate at least 50 percent higher than the old income tax to raise the same revenue (or a much larger deficit, if you wanted to avoid imposing a tax increase on most workers). Exempting a poverty-level income from taxation through increased standard deductions, personal exemptions, or tax credits can prevent a tax increase on lowest-income taxpayers, but about twice as many middle- and upper-income taxpayers would get a tax hike as a tax cut.
It thus didn't take a crystal ball when I warned my former boss Jack Kemp on the first day of hearings by the 1995 GOP National Commission on Economic Growth and Tax Reform, which he chaired, that "all the 'consumption' taxes are going to self-destruct politically, when it becomes apparent that any revenue-neutral version will represent a tax increase for working families." And that's exactly what happened to the Armey-Forbes plan in the 1996 Republican presidential primaries--unfortunately, not before the party commission had endorsed the politically suicidal approach.
The fairest, simplest, and most efficient way to reform the tax code would involve not just treating labor and property income alike, but also simplifying the means of collection, so that taxes are filed mostly by employers (business, government, nonprofit, or the self-employed) rather than by individuals. This would be a radical simplification for the tens of millions of people who would no longer have a personal relationship with the IRS every April, but it is a less radical proposal than it may at first sound. It would simply treat all income in the same way that President Bush proposed last year for dividends: non-deductible when paid by businesses, but non-taxable when received by individuals.
Recall the example of the family purchasing a new car or computer. The business pays out the money entirely as income to its employees, investors, and creditors. Under our current system, the business deducts all that compensation (except some dividends) from its taxable income, and the government taxes the recipients' income. If instead the compensation paid by the business (or other employer) were non-deductible, the existing corporate and personal income taxes would be superfluous, because the tax on all that labor and property income would already have been "prepaid." There would, however, be a per-person rebate for the income and payroll taxes paid by the employer on income below the poverty line ("human maintenance").
If a flat tax rate were applied to such a broadened tax base, there would be three interesting results. First, the tax code's complexity would disappear. Most families would have no contact with the IRS except to receive their rebate. Instead of the IRS having to collect tax from more than 100 million taxpaying entities, it would have to track only a few million businesses. Second, both the tax code and the economy would be vastly more efficient. The same amount of total revenue as now raised by both the current corporate and personal income tax codes could be raised with a flat rate of about 16 percent, less than half the current code's top rate. To balance non-Social Security federal spending and eliminate the current deficit, a rate of 18 percent would be required. There would be no double taxation of any income, or any incentive to make decisions for tax reasons other than economic efficiency. Third, the tax burden under such a system would be about as "progressive" as the current tax code. This is because even though both corporate and personal income taxes with their progressive tax rates would be replaced, all property income now excluded from the tax base would be taxed, while the sub-poverty line "human maintenance" costs contained in labor compensation would not be taxed.
The same economic facts of life also explain the economic and political logic of Social Security reform. While income varies according to the lifetime pattern described above, we all need to be fed, clothed, sheltered, and transported, whether or not we earn income. Our income typically exceeds consumption during parenthood and the "empty nest," while consumption exceeds income during childhood and retirement. This last fact creates what might be called the "retirement gap." When people retire, labor compensation falls to zero, yet consumption is usually much higher than the property income from earlier saving. The basic problem is how to fill this gap without forgoing retirement, suffering a sharp fall in consumption during retirement, or lowering total lifetime earnings and consumption (which is what happens if early in life one invests more in lower-yielding property and less in higher-yielding human capital).
Without government, the retirement gap can be bridged only by love--a gift from someone (most often one's adult children) whose own consumption is thereby reduced. Pay-as-you-go Social Security solved the retirement problem by providing an asset that the private financial markets cannot. While a financial account is essentially a claim on property, a pay-as-you-go Social Security retirement pension amounts to a share in a diversified human capital mutual fund. Social Security makes it possible, in effect, for workers, by pooling a fraction of their income, to transfer labor compensation from their working years to retirement, and to surviving dependents after their deaths. However, once pay-as-you-go benefits have closed the retirement gap, any further expansion of benefits necessarily comes at the expense of smaller investment in either children or productive property. The pay-as-you-go system therefore requires periodic readjustment to maintain proper balance among these three competing imperatives. Reagan and Congress undertook such an adjustment in 1983. "The press is trying to paint me as now trying to undo the New Deal," Reagan complained in his diary in January 1982. "I remind them I voted for FDR four times. I'm trying to undo the 'Great Society.' It was LBJ's war on poverty that led us to our present mess."
Today, any Social Security reform plan must start with a simple fact: Social Security's expected future deficits are entirely the result of lower birthrates (owing in part to three decades and counting of legal abortion). If that trend were reversed, the deficits would be easily surmounted. But let's assume it won't be. Options for filling the retirement gap are still much broader than Republicans have so far considered. What's clear is that plans to replace pay-as-you-go Social Security with compulsory individual financial accounts are based on the same Stork Theory as the consumption tax, and would worsen the demographic problem. As the redoubtable Martin Feldstein, the GOP's preeminent "stork theorist," has explained, "The essential feature of the transition to a funded program of retirement benefits is a period of reduced consumption by employees during the early years of the transition so that a dedicated capital stock can be accumulated. This dedicated capital is then used to finance retirement benefits, thereby permitting lower taxes and more consumption by employees in later years."
Note that in Feldstein's view, "capital" means "nonhuman capital": education, training, and child-rearing are classed as "consumption," not investment in human capital. The analysis takes the population and its skills as given, and only by virtue of this false assumption comes to the equally false conclusion that taxes on property income not only should but inevitably must be shifted to workers. As the demographic implosions in Europe and Japan suggest, heavily shifting to "consumption" taxes causes a sharp decline in investment in human capital.
When the pay-as-you-go system started, its rate of benefit return on payroll tax contributions was much higher than the stock market, because there were many workers supporting few retirees; but in the long run the average rate of return on pay-as-you-go Social Security is equal to the rate of economic growth (which over the past 75 years has been higher than the average yield on government bonds). Replacing pay-as-you-go pensions with individual financial accounts, as Feldstein forthrightly acknowledges, requires the startup process to go into reverse: The first generation to work under the new regime must continue paying the benefits for their parents' generation while saving for their own retirement.
The size of this transition tax was strikingly illustrated when the author of the leading plan to divert Social Security taxes into private accounts recently had the plan costed out by Steve Goss, Social Security's chief actuary. Until then, its proponents had argued, with apparent plausibility, that the transition cost might be financed by borrowing several trillion dollars, halving discretionary spending, and reaping increased tax revenues from increased investment in plant and equipment. But Goss's estimates made three unpleasant features of the plan explicit: First, general-revenue "transfers to the Trust Fund would, however, not be contingent on achieving these [cuts] in actual federal spending"; second, there would be no net increase in plant and equipment, economic growth, or revenue "feedback" if the cost were financed by federal borrowing; and third, replacing the payroll taxes diverted from the Trust Fund without additional borrowing would require tax increases, mostly on individual and corporate income, lasting several decades and ranging up to more than 7 percent of taxable payroll.
PROPONENTS, armed with millions in lobbying dollars but few popular votes, are assuming that this plan will be endorsed by President Bush, brought by Republicans to the House and Senate floors, and adopted by majorities in both chambers. I think it highly unlikely that any one of these will happen, let alone all three. I don't see how President Bush could possibly endorse such a plan, because it violates his two basic principles for Social Security reform: that any accounts must be voluntary, and that they must involve no tax increase. The private accounts aren't voluntary, because if you forgo the private account option, you must still pay the government's cost of funding everyone else's account. And the plan would mean a much larger and earlier tax increase than even John Kerry proposed.
I don't think it will happen, further, because White House strategist Karl Rove wants to replay the successes, not the mistakes, that followed William McKinley's 1896 and 1900 campaigns on the Lincolnian slogans "Honest Money" and "A Full Dinner-Pail." Upon reelection in 1900, McKinley announced that he was reversing his support for a high tariff--the consumption tax of its day--and would aggressively apply antitrust law. Theodore Roosevelt, who succeeded McKinley upon the latter's assassination, followed this strategy, and ultimately endorsed an income tax to replace the tariff. Roosevelt's differences with his anointed successor, William Howard Taft, precisely over the tariff-versus-income-tax issue split the Republican party in 1912, handing the presidency to Woodrow Wilson with only 43 percent of the vote, and giving control of Congress to the Democrats.
As for the Republican Congress, I don't believe it will like the plan any better than it did the Feldstein plan of 1999, which died in Republican-controlled committees, when dismayed GOP lawmakers discovered the "essential feature": a 75 percent "clawback" tax on all money withdrawn from personal retirement accounts, after which the plan still wasn't fiscally balanced. They'll have a similar shock when they discover the huge income-tax "clawback" in the new Ferrara-Ryan-Sununu version. The Goss memo amounts to opposition research for all 2006 challengers to any Republicans who cosponsor or vote for the plan.
The Social Security reform that makes most sense to me is a variation of a plan long advocated by former Republican Social Security chief actuary Robert Myers. It would cut payroll taxes immediately, thus getting rid of the Social Security "surplus," which Congress has simply been using to fund deficits in the rest of the federal budget. In exchange, there would be a matching reduction in the level of future retirement benefits (prorated for the share of working years that each worker received the tax cut), eliminating expected future deficits. I proposed an "80 percent" solution for current payroll tax rates and future benefits to the Kemp commission, but after nine more years of wasted excess payroll taxes, I would now recommend a "75 percent solution." This plan is truly voluntary, because families would be free but not required to put the payroll tax-cut money into financial retirement accounts; for most families with children, education would be a more pressing (and higher-yielding) investment.
Since the plan reduces current payroll taxes and future benefits in the same proportion, the rate of return per dollar of payroll contribution would be higher than under any of the other alternatives, which involve tax increases and benefit cuts. The plan also depends far less than the other plans on the accuracy of forecasts for economic growth and financial asset returns. If economic growth is as slow in the future as the actuaries have been projecting, then families will be prepared and the system will be kept in balance with relatively small adjustments. But if economic growth continues to outperform the actuaries' projections, as it has in recent years, the system will be in surplus and payroll taxes can be cut again, the benefit reductions halted, or both.
In short, a simple, low, flat-rate income tax modeled on President Bush's dividend proposal, combined with a truly voluntary plan to balance the pay-as-you-go Social Security pension system without income-tax funding, remains eminently doable. And it is the way to continue the just, efficient, and popular fiscal legacy linking Abraham Lincoln and Ronald Reagan, and finally realize Reagan's vision of not "a temporary uneasy alliance, but the creation of a new, lasting majority."
John D. Mueller is a financial-market economist in Washington, D.C. He was staff economist to then-congressman Jack Kemp from 1979-88 and in 1995-96 served as an adviser to the GOP's National Commission on Economic Growth and Tax Reform.