In 2004 the American Prospect staged a website "debate" over whether President Clinton's domestic policy had been sufficiently liberal. Ann Lewis insisted that Clinton's presidency had been "good for our country," an argument the former communications director in the Clinton White House could have made under sedation. Disputing her claim was Max Sawicky, an economist at the Economic Policy Institute, a liberal think tank. His central accusation was that Clinton had failed to "rehabilitate the reputation of the welfare state by proposing well-founded expansions. This, I submit, is the mission of the Democratic Party; otherwise, it has little purpose." He went on to explain that there is "no mystery about how to fill" the "huge gaps in our social safety net." You do it with social insurance.

If the expansion of the welfare state is the reason liberals get up and go to work in the morning, its contraction is the reason conservatives do. Almost any page from the writings of Ronald Reagan will demonstrate this point. To pick just one example, Reagan told the American Bar Association in 1983, "It's time to bury the myth that bigger government brings more opportunity and compassion…. In the name of fairness, let's stop trying to plunder family budgets with higher taxes, and start controlling the real problem—Federal spending."

This argument—over the proper size of America's welfare state—has been going on for 75 years. Three things might prevent it from going on another 75, but two of them are unlikely. The first is that one side will score a decisive victory over the other, winning (so to speak) all the arguments and all the elections. The second is that the two sides will split the difference in a way they both feel reasonably happy about. The third, less far-fetched possibility, is that the debate will not be resolved but abandoned—after political and intellectual exhaustion motivates the combatants to redefine what they're arguing about.

Growth of the Welfare State

Before we can speculate about the future of the welfare state controversy, we need to understand where we are now and how we got here. In his indictment, Sawicky pronounces the growth rate of the welfare state during Clinton's administration "pathetic." He arrives at that conclusion, however, by using a dubious measurement: discretionary federal spending on everything except national defense. This category, used by the Office of Management and Budget (OMB) in its historical tables accompanying each year's federal budget, includes government endeavors no one thinks of as components of the welfare state, such as programs for international affairs, science, space, and technology. At the same time, it excludes the social insurance programs Sawicky says are vital: Social Security, Medicare, Medicaid, and food stamps, among others.

A better yardstick can be fashioned to keep score of the Hundred Years' War between the welfare state's expanders and its reducers. One of OMB's historical tables tracks annual federal outlays beginning in 1940, using enormous categories ("superfunctions") composed of smaller but still vast "functions." The two main superfunctions, National Defense and Human Resources, have together accounted for at least 61.7% of federal outlays in every year since 1940. The superfunctions that account for the rest of federal outlays are:
 

• Physical Resources (e.g., Energy, the Environment)
• Net Interest on the National Debt
• Other Functions (e.g., Science, International Affairs, Agriculture, General Government, and the Administration of Justice)

If we group those three final superfunctions as "Everything Else," we can see the changing makeup of the federal budget over the past 67 years in Chart A.

OMB's Human Resources superfunction is made up of the following six functions:

• Education, Training, Employment, and Social Services
• Health (excluding Medicare)
• Medicare
• Income Security (excluding Social Security)
• Social Security
• Veterans' Benefits and Services

Let's leave the final category off the list, since veterans' programs don't figure prominently in the arguments liberals and conservatives have conducted over the size and scope of government. What's left is an imperfect but useable approximation of the welfare state. (The chief imperfections are that we're looking only at federal expenditures, which leaves out state and local welfare state spending; and that the costs imposed by regulations, like the minimum wage laws, don't show up in the federal government's outlays.)

The OMB data comes with a "composite deflator," which allows us to remove the distorting effects of inflation by expressing every nominal dollar figure its tables capture in terms of a constant—the dollar's value during the fiscal year 2000. We can also adjust for population growth—there were 2.28 times as many Americans in 2007 as in 1940—by dividing the annual, inflation-adjusted outlays for Human Resources by the Census Bureau's annual population estimates. Doing so yields Table A.

This table reveals that the welfare state battle between liberals and conservatives has been as evenly matched as the one at Little Big Horn between Sitting Bull and Custer. Real, per capita federal spending on Human Resources was 15 times greater in 2007 than in 1940. Whatever else it may tell us, this 1,394% increase is one more demonstration of the power of compound interest. You achieve that huge expansion over 67 years with an annual growth rate of 4.10%, which doesn't sound so formidable.

The 67 years in question have seen, of course, varying growth rates for per capita Human Resources outlays, not a constant one. On 11 occasions since 1940 these outlays declined from the previous year; on 14 others they increased by more than 10 percent. Table B segments the growth of the welfare state into presidential terms and presidencies. (Note that although presidents are inaugurated in the midst of a fiscal year, their budget proposals, and the decisions made by Congress about spending, taxing, and borrowing, are almost always directed to the ensuing rather than the current fiscal year. For example, President Carter took office in 1977, but the budgets he controlled were for 1978, 1979, 1980, and 1981. The 1977 budget had been determined by Congress and President Ford in 1976. This chart takes Human Resources outlays in FY 1977 as the baseline for measuring spending under Carter.)

The discovery that real, per capita Human Resources outlays expanded seven times as rapidly under President Eisenhower as under President Clinton is decidedly counterintuitive. We need to place that fact in context. The chart shows that the general trend has been for the growth rate of the American welfare state to decline. Conservatism's political advances account for some of this slowdown. But much of it is better explained by a generalization more often associated with the life sciences than the social sciences: immature entities grow more rapidly than mature ones. It's normal and healthy for a child's weight to double between the first and fourth birthdays, but also to increase by only a third between the tenth and thirteenth birthdays. (Additionally, real, per capita Human Resources outlays declined by 61% between 1940 and 1945; Dr. Win-the-War replaced Dr. New Deal, in FDR's phrase. It took nearly the entirety of Truman's term in office for those outlays to climb back to where they were before the war, making the 1940s a lost decade for liberals' ambitions to increase the welfare state.) 

The welfare state's growth in the 15 years after World War II was driven by the maturation of the Social Security program, which issued its first benefit check in 1940 and grew (adjusting for inflation and the changing size of the national population) 27% a year between 1945 and 1953. Social Security outlays accounted for only 15% of Human Resources spending in 1945, but 37% in 1953. This trend continued under Eisenhower; Social Security outlays increased at an annual rate of 15%. By 1961, Social Security accounted for 52% of Human Resources outlays (leaving aside, as we have in all these calculations, outlays for veterans' programs).

Important components of the welfare state were added, one-by-one, as it matured. Congress made disability insurance part of Social Security in 1956. It created a number of new programs in 1965, including Medicare, Medicaid, the Elementary and Secondary Education Act, and the Higher Education Act. Because of such Great Society programs, real, per capita federal outlays for Education, Training, Employment, and Social Services grew at an annual rate of 31% between 1965 and 1969, three times as fast as they grew under Truman and five times as fast as under Eisenhower. Outlays for medical programs, other than Medicare, grew at an annual rate of 24% during those years, which was, once again, much faster than the rate before 1965.

Medicare is a separate, special category. Since dispensing its first dollar of benefits in 1966, Medicare has grown at an annual rate of 16.9%, after adjusting for inflation and national population growth. Measured in constant dollars, it accounted for 16% of all Human Resources spending between 1966 and 2007. The most significant growth in Human Resources spending is attributable to Medicare and "Health Care Services," an OMB category dominated by Medicaid. Still using constant dollars, these two categories combined to account for 8% of Human Resources outlays under Kennedy and Johnson, 15% under Nixon and Ford, 17% under Carter, 21% under Reagan, 26% under George H.W. Bush, 31% under Clinton, and 34% under George W. Bush. Measure all the Human Services outlays from 1962 (the first year of more detailed OMB historical tables) through 2007 in constant dollars, and it turns out that Medicaid, Medicare, and Social Security accounted for just under two-thirds of the total.

The Persistence of Poverty

The long political advance of liberalism has coincided with the refusal of any prominent liberal politician or writer to specify or even suggest the welfare state's ultimate and sufficient size. Instead, liberals have denounced our shockingly insufficient welfare state every year since the beginning of the Progressive era. When Max Sawicky did so in 2004, real, per capita expenditures on Human Resources were more than twice as large as they had been in 1975.

Yet it would be absurd to argue that the sort of economic insecurities the welfare state exists to alleviate were twice as severe in 2004 as in 1975, or that America had been little better than a Third World country during Gerald Ford's presidency. The percentage of Americans who owned their own homes increased between 1975 and 2004 from 64.4% to 69.1%. Average life expectancy became 5.2 years longer. In 1975 the proportion of Americans aged 25 or older whose educational attainments included the completion of at least four years of high school was 62.5%, and 13.9% had completed at least four years of college. By 2004 the percentages were 85.2% and 27.7%, respectively. 

Numerous consumption items that had been luxuries or Research-and-Development daydreams in 1975 were parts of the furniture of American life in 2004, even for millions of Americans with incomes below the median: e.g., color televisions receiving dozens of channels by cable or satellite, home computers accessing the internet, air conditioning in homes and cars, cell phones and microwave ovens. None of these developments give pause to liberals who say that a welfare state that doubles its outlays in the 29 years separating a prosperous era from an even more prosperous era needs to grow dramatically faster.

Nor do liberals ask hard questions about how the persistence of shocking and shameful poverty relates to this inexorable growth of the welfare state. Americans were jarringly reintroduced to their economically vulnerable and socially isolated countrymen by Hurricane Katrina. Ask any hopeful Democrat leaving an Obama rally what we should do about such poverty, and you'll be told that the federal government ought to spend a lot more money to help these people. What you won't be told is that a welfare state that grows 4% a year for six decades and still hasn't eliminated the nation's worst poverty might have problems that more money can't solve. Specifically, you won't hear serious consideration of the possibility that the benefits already dispensed by the welfare state are not so much scandalously inadequate as scandalously misallocated. 

Under the right economic circumstances liberals could afford to be blithe about devoting many of the welfare state's dollars to purposes that are less than urgent. The years of the welfare state's fastest growth were also, not coincidentally, the years of America's postwar economic boom. During the quarter-century after 1945, America competed from a position of unprecedented strength against nations whose industrial bases and civilian populations were decimated by the war. This uniquely favored economy threw off a reliable "fiscal dividend" that raised government's tax revenues in the absence of tax increases. World War II's frontline nations have long since reindustrialized, however, and been joined by growing numbers of newly industrialized countries. America may well experience strong economic growth in the future, but is unlikely ever again to enjoy the latitude to hand out generous social benefits without looking at the price tags.

Under the right political circumstances, liberals could ignore these economic realities, pretend we're still living in John Kenneth Galbraith's "Affluent Society," and party like it's 1959. The Reagan legacy, however, is certain to thwart them. The liberal historian Sean Wilentz has recently argued that "never, in our lifetimes," are we going to see top marginal income tax rates of 70% again, something that "if you stood looking to the future in 1980, would have been amazing." 

Hunting for Votes and Revenues

The liberals' problem is two-fold. First, they don't have the political latitude to soak the rich. Second, they've taken the position that hardly anyone is rich. In a debate before the Pennsylvania primary Hillary Clinton said, "I am absolutely committed to not raising a single tax on middle-class Americans, people making less than $250,000 a year." Barack Obama's reply at the time was less categorical, but on the eve of the Democratic convention his top two economic advisors signaled, in a Wall Street Journal op-ed, that the Democratic nominee had embraced the Clinton position with one small modification: the Obama tax plan "would not raise any taxes on couples making less than $250,000 a year, nor on any single person with income under $200,000—not income taxes, capital gains taxes, dividend or payroll taxes." 

The difficulty, as the American Prospect's Ezra Klein put it delicately, is that "when everyone below the 95th percentile is untouchable, and effectively middle class, you're in a bit of an odd discourse, distribution-wise." Strictly speaking, Clinton and Obama have made everyone below the 98th percentile untouchable, since only 1.92% of American households had an income of $250,000 or more in 2007. (The portion of all households with an income of $200,000 or more is 3.64%.) 

There is no evident escape from this odd discourse. Democrats need to hunt where the ducks are. The dilemma is that they're hunting both for votes and for tax revenues. Their electoral strategies require votes (and campaign contributions) from the 30.1% of households with an income between $75,000 and $250,000. (Given the positive correlation between household income and voting turnout, these families make up significantly more than 30.1% of the electorate.) Democrats' welfare state aspirations, however, cannot be realized merely by making the rich less rich. Since even a 70% tax bracket won't do much good if it applies only to one-fiftieth of the population, the revenue yield from a much lower, politically feasible top bracket is going to be underwhelming. The math is unyielding: enacting any significant portion of the liberal agenda will also require making the merely comfortable noticeably less comfortable—and Democrats are terrified that advocating or enacting tax increases on upper middle-class voters will doom them when those voters go to the polls.

Even explicit pledges to soak the rich, and only the rich, do not guarantee electoral success with voters who aren't. In 2006, for example, the film director Rob Reiner promoted a ballot proposition in California to guarantee that every 4-year-old in the state would attend preschool. It called for dedicating to the preschool program the projected $2.4 billion per year from a new top bracket for the state income tax—11% as opposed to the existing top rate of 9.3%—affecting couples making more than $800,000 per year and individuals making more than $400,000. 

Reiner was optimistic in advance of the vote. The "American public has no problem raising taxes if it's for something good," he told the columnist E.J. Dionne, who expressed the hope that "California, which started the tax revolt in the late 1970s, could inaugurate a new era of public investment in things that matter." After the ballot proposition lost by a 61–39 margin, Dionne admitted that "there remains a deep skepticism about government spending, even for the best purposes…. Attacks on tax and spend sound old and tired, but they still have force."

The Problem of Entitlements

The welfare state grew less during the eight years of the Reagan Administration than during a typical year in the glory days of the Great Society from 1965 to 1973. In the context of conservatism's long losing streak, this is an exceptional achievement. And yet…while the 0.93% annual growth rate of real, per capita Human Resources outlays from 1981 to 1989 is a small number, it's also a positive number. If this is the best conservatives can do, then the question is not whether liberals or conservatives will win the battle over the welfare state, but whether liberals will win sooner or later.

The central problem, both in terms of mathematics and politics, is entitlements. The Reagan Administration made good on its promise to curb the size of the federal establishment—in those areas where it was politically feasible to do so. Between 1981 and 1989 real, per capita federal outlays on Education, Training, Employment, and Social Services programs shrank at an annual rate of 4%. Income Security outlays declined 0.9% per year. But these two functions, together, accounted for only 34% of the constant-dollar spending on Human Resources during the Reagan years. By contrast, Social Security, Medicare, and Health Care Services, together, accounted for 64.4% of Human Resources outlays, and grew at an annual rate of 2.4%—not dramatic but, under the circumstances, relentless.

Reagan's first OMB director, David Stockman, wrote in his memoirs that the Reagan tax cuts could "add up only if deep dents were kicked in the side of the welfare state," and some of those kicks had to be directed at entitlements. "No single issue was as critical to the success of the Reagan Revolution as Social Security reform." The Reagan Administration did lean against Social Security, and the results were not negligible. Modest and gradual increases in the retirement age were enacted in 1983. That year Congress also subjected, for the first time, a portion of Social Security benefits to the income tax, which doesn't show up on Social Security's books as a benefit reduction but amounts to backdoor means testing.

All the Reagan Administration's more aggressive efforts at entitlement reform, however, were soundly defeated and quickly abandoned. In May 1981 the White House floated Stockman's reform plan, which included steep reductions in the pension benefits paid to retirees who begin collecting at age 62, a sizeable majority of Social Security recipients. The Republican-controlled Senate quickly passed, unanimously, a resolution denouncing the plan, and Reagan spoke of it in a televised speech in September as though some other president's administration had come up with this hare-brained scheme.

The Gingrich Republicans who swept to congressional majorities in 1994 deserve much of the credit for the "pathetic" record Max Sawicky scorned: Human Resources outlays grew more slowly under Clinton than any president except Reagan. Those conservatives, however, were also thwarted by politically impregnable entitlements. Clinton repeatedly "used Medicare's popularity against Republicans," according to Major Garrett's book on the Contract with America, The Enduring Revolution (2005). Republicans insisted that their plan would have cut the rate of Medicare's growth, not actually reduce the program itself, going so far as include in their 1996 platform the raw statement that "Bill Clinton lied about the condition of Medicare and lied about our attempts to save it." None of these efforts did them or their plans for Medicare any good. By the end of Clinton's presidency the Republicans were "demoralized, frustrated, and listless," according to Garrett. "They wanted to change Medicare but they didn't know how."

Both of these futile assaults against entitlement programs had antecedents. Neither in 1980 nor 1994 did Republicans seek or secure a mandate to kick deep dents in the side of the welfare state. In his acceptance speech at the 1980 Detroit convention Ronald Reagan promised a federal hiring freeze and a thorough review of every federal program to eliminate waste and inefficiency. He promised to turn over to the states and localities every federal program that they could run more effectively, along with the funding sources to pay for each program. But that was about it, and that was all pretty general.

The 1994 Contract with America was equally reticent on the subject of shrinking government. Of its ten provisions, the only one that explicitly sought a smaller welfare state called for a "personal responsibility act" to reduce welfare eligibility and benefits, a promise the Republican Congress fulfilled in 1996. Another provision called for a constitutional amendment to require a balanced budget, and for a line-item veto, a measure likely necessitating a constitutional amendment as well. The purpose of these mechanisms, according to the Contract, was to "restore fiscal responsibility," but both were procedural changes that did not commit the Republicans to any specific spending cuts or ask the voters to endorse any. The other eight provisions addressed disparate issues, including crime, national defense, term limits, and the internal management of Congress. 

Not by happenstance did Republicans campaign against the welfare state at a high level of generality, avoiding specific pledges to scale back popular entitlement programs. The professionals who ran the 1980 and 1994 campaigns understood that their candidates could not drag such anvils behind them and still win the race. Nor was it by accident that Republicans, having secured power on such strategically fashioned platforms, found it impossible to launch successful post-election campaigns for reducing entitlements by uttering sentences that began with the words, "Oh, by the way."

As we've noted, however, liberals are in a box, too, since Democrats dread the thought of launching a post-election campaign for upper middle-class tax increases with the words, "Oh, by the way." The result is a paradox. The contest between liberals who want to expand the welfare state and conservatives who want to shrink it would seem to be a classic zero-sum game, a tug-of-war in which one side gains precisely to the extent the other loses. Yet both speak as if it's obvious that the other is running, and ruining, the country. Either could be wrong, but both can't be right.

Era of Bad Feelings

One explanation for this paradox is that each side's activists and theoreticians are playing a long game, looking past the stubborn political realities that elected officials are not at liberty to disregard. Conservative activists and theoreticians will not relinquish the hope that the policy consequences of the 1936 election might yet be reversed, and the New Deal undone. They are as angry at conservative politicians, whom they consider too unprincipled and maladroit to advance this goal, as they are at liberals. Liberal activists and theoreticians will not relinquish the hope that the policy consequences of the 1984 election might yet be reversed, and the Glorious March Forward resumed as though the Reagan detour had never happened. They are, by the same token, contemptuous of liberal office-holders who accommodate the Reagan legacy rather than work day and night to exorcise it. 

In both cases, the activists and theoreticians, many of whom are tenured, make the willingness to run suicide missions the sole criterion for judging the courage of politicians, who are never more than an election cycle away from having no votes to cast and no policies to shape. Those who don't survive the electoral short game never get to play the policy long game, of course. No conservative, either in the trenches or the commentariat, has yet devised a strategy for politicians to kick deep dents in the side of the middle-class entitlement programs without forfeiting a presidency or a congressional majority. Similarly, liberal writers and activists are much better at coming up with ways for liberal politicians to spend additional public funds than with ways for them to raise the taxes needed to acquire those funds without losing office.

This stalemate, where conservatives can stymie liberals' ambitions and liberals can stymie conservatives', is one of the reasons prosperity and peaceful victory in the Cold War have bequeathed our long Era of Bad Feelings. The stalemate is not tenable, although the bad feelings are likely to persist. The oldest baby boomers, born in 1946, are turning 62 in 2008, and beginning to collect early Social Security pension benefits. In 2011 they'll be eligible for Medicare and full pension benefits, if they chose to defer their retirement. American birth rates didn't peak until 1957, so the ranks of beneficiaries will grow for many years. The Census Bureau projects that 40.2 million Americans will be 65 or older in 2010, representing 13% of the entire population. By 2025 there will be half-again as many people 65 or older, 63.5 million, amounting to 18.2% of the population.

The baby boomers' retirement will be the best documented, least surprising policy challenge in American history—and still we are not prepared for it. Herb Stein's Law remains operative, however: if something can't go on forever, it won't. Entitlements can't go on, indefinitely, laying claim to a bigger portion of the federal budget and the GDP. Once the furniture is engulfed in flames we will finally start shopping for fire extinguishers. 

Hard Choices

If conservatism's batting average against the welfare state is predictive, the fire drill will involve benefit reductions and tax increases…but mostly tax increases. This result sounds like liberals' next victory in their zero-sum game with conservatives. Human Resources outlays will increase, and the welfare state will lay claim to a bigger slice of GDP. Whether liberals will and should feel good about this victory is a separate question, however. If a "new era of public investment in things that matter" will at last vindicate and resume the New Deal/Great Society agenda, liberals are going to have to sort out whether all the things that matter matter with equal urgency. If so—if Human Resources expenditures are indistinguishable, interchangeable units—then the federal government's additional pension and medical benefits payments to retired baby boomers, many of whom are otherwise prosperous, will be an entirely satisfactory result for liberals. 

And if not, then not. The journalist Matt Miller has tried, with little success, to get liberals to fashion policy on the basis of rigor rather than sentiment and wishful thinking. He notes that Medicare and Medicaid are scheduled to absorb 9% of GDP by 2030, twice the amount they do now. "Then there's the plethora of other Democratic priorities, from covering the uninsured, to wage and child-care subsidies for the working poor, to R&D and infrastructure backlogs." All the leading Democrats want to repeal the Bush tax cuts, as they apply to households making more than $250,000 per year. The trouble, according to Miller, is that you can repeal them only once. 

Coming up with six new ways for the government to spend each additional dollar of revenue may be good politics but cannot be good governance. However, asking liberals to acknowledge that there can be only 100% of anything, including GDP at the disposal of the welfare state, and thus requiring them to think hard about domestic policy trade-offs, is like trying to row a canoe up a waterfall. The political (and financial) capital liberals consume with tax increases that shore up our existing entitlement programs is capital they can't consume again with the additional tax increases needed to expand child-care, develop green technology, protect workers from the consequences of globalization, etc. How do we render entitlements solvent, and pay for liberals' numerous initiatives for those who aren't elderly, and do all this without resorting to the kind of tax increases that imperil Democrats and the economy? Miller posed that question to one Clinton Administration veteran, who said, "I don't think that conversation has yet taken place in the heads of most Democratic economists." 

It would be a good conversation to have, not just for Democratic economists, but for the whole country. Each of our two ideological adversaries has a firm grip on one end of the domestic policy wishbone. Liberals speak to and for the public's aversion to a smaller welfare state. Conservatives speak to and for the public's aversion to higher taxes. It's not possible for them to go on pulling in opposite directions indefinitely without guaranteeing bad luck for everyone.

Neither side, however, wants to relinquish its hold on that part of public opinion that is the source of its own political strength. Both hope that the tension inherent in the public's desire for a welfare state that confers generous benefits while imposing modest taxes will ultimately be resolved in its own favor. Three years ago the New Republic's Jonathan Cohn argued for the obvious liberal resolution: We "could actually raise taxes," he said, since "there's a compelling case for the government spending a great deal more money than it does now." Like Rob Reiner and E.J. Dionne, Cohn was optimistic that the public might be receptive to this compelling case. The "early signs suggest that Americans might be more hospitable to tax increases than everybody has long assumed," he wrote.

He sounds less hopeful today, despite the Democrats' victories in the 2006 elections and favorable prospects in 2008. Even if Senator Obama wins the presidency, Cohn says, it's not clear "that he can restore public faith in liberal ideals." Such a restoration would take us back, politically, to the middle-third of the last century, when the New Deal's popularity sustained liberalism and most Americans "believed the government was capable of doing the right thing." After "the backlash against Democratic policies on race, taxes, and national security" that began in the 1960s, however, America "returned to its natural predisposition: an instinctive skepticism of the public sector."

Cohn describes conservatives' efforts to exploit this skepticism in apocalyptic terms. They have "declared war on the welfare state," he wrote last year, so liberals "should probably defend said welfare state's existence before harping on its modest, if still regrettable, flaws." Conservatives did declare war on the welfare state in 1936 and 1964, but in the 26 fiscal years since Reagan's inauguration in 1981 the federal government's real, per capita Human Resources outlays grew by 74%. Whatever declaration of war Cohn is referring to, it would be very hard to find evidence that modern conservatives have waged war on the welfare state.

Supply-side economics was supposed to be a way for conservatives to avoid a fight they didn't think they could win. Shortly after Reagan's inauguration in 1981 National Review editorialized in favor of the "political genius" of the Kemp-Roth tax cuts, calling them a no-lose proposition for conservatism. Enact the cuts, NR said, and then
 

either, or a mixture, of two things happens: a) an economy stimulated by the tax cuts pours revenue into the federal coffers, making subsequent budget cuts milder and politically more palatable; or b) that doesn't happen, and sudden inflationary pressures make deep cuts to reduce the deficit much more feasible politically.

Whether, or to what extent, the economic proposition that supply-side tax cuts would pour revenues into federal coffers was vindicated remains a debated proposition. It would be much harder to argue, however, that the political part of NR's prediction has held up well. The most favorable case one could make for it is that the welfare state grew more slowly than it would have in the absence of the tax cuts. The problem is not so much that this claim, whether or not plausible, is unprovable. It's that conservatives who expected that tax cuts would determine whether the welfare state's reductionwould be large or small aren't going to derive much solace from a debate about the extent to which tax cuts determined whether the welfare state's expansion was fast or slow.

Supply-side tax cuts did little to necessitate or even facilitate reducing the welfare state, and there is no reason to believe an explicit campaign for that goal will succeed where Barry Goldwater's failed. Given all that, conservatives need to weigh the costs and benefits of putting liberals' minds at ease by explicitly renouncing the war against the welfare state, the one that's barely being waged and steadily being lost. They could do so by making clear that America will and should have a welfare state, and that the withering away of the welfare state is not the goal of the conservative project, not even in the distant future. What libertarians will regard as a capitulation to statism is better understood as conceding ground conservatives have been losing for 75 years and have no imaginable prospect of regaining.

The political advantage of this concession is that it leaves conservatives positioned to argue for a better, smarter, and fairer welfare state. "Liberalism needs government," says Cohn, "because government is how the people, acting together, provide for the safety and well-being of their most vulnerable members." Very well, but in a society that is remarkably prosperous by global and historical standards, shouldn't "most vulnerable members" be construed as referring to the most vulnerable 5, 10, or 25% of the population—not just the abjectly miserable, let us concede, but people confronting serious threats or problems? Yet when it turns out, time and again, that the effective meaning of liberal welfare and social insurance programs is to elicit compassion and government subventions for the most "vulnerable" 75, 80, or 95% of the population, it's hard not to feel scammed.

Good Government

If they were to accept conservatives' stipulation that the welfare state war is over, liberals could turn their attention from defending its existence to examining its flaws, which are definitely regrettable but not necessarily modest. The distribution of the welfare state's benefits leads the list of those flaws. Like Cohn, Paul Starr of Princeton University and the American Prospect, says the welfare state is about the poor. Its "objective should be, above all, to eliminate poverty and maintain a minimum floor of decency to enable individuals to carry out their own life plans." But giving benefits to everyone, not just the most vulnerable, serves social and political purposes. Socially, "the long-term tasks of nation-building and of fostering a common culture and a sense of shared citizenship also strongly argue for public and universal schooling, old-age pensions, and other services that serve an integrative as well as egalitarian purpose," according to Starr. Politically, the imperative to construct democratic majorities that support programs for the poor "will often mean support for programs that provide universal benefits." We may say that such programs "target" the most vulnerable 100% of the population.

Neither argument excuses liberals' refusal to concentrate the welfare state's benefits on the most vulnerable, whose plight is ostensibly liberalism's animating concern. There is much to be said for public education; some of it is said by Democrats—such as the Clintons, Obamas, and Bidens—who enroll their children in private schools with no apprehensions that they will be warped into selfish Republicans by the time they graduate. The social cohesion promoted by paying taxes to and receiving benefit checks from the same government agency as millions of other strangers is spurious, however—an example of the "granfalloon," Kurt Vonnegut's term for "a proud and meaningless association of human beings."

Equally spurious is the supposed necessity of universalizing social insurance in order to secure majority support for programs that only a minority really needs. Were it otherwise, the Earned Income Tax Credit (EITC), for which all families with an income above $41,646 are ineligible, would be as politically besieged as Aid to Families with Dependent Children was before 1996. In reality, EITC enjoys bipartisan support in Congress and is politically unassailable. "If means testing makes Social Security as unpopular as the EITC," writes Mickey Kaus, "Democrats have nothing to fear."

Liberals, in short, should take Yes for an answer. 75 years of their rhetoric about defending the most vulnerable among us really has persuaded the American people, who are fully prepared to support, on the merits, government programs to help the needy. For everyone else—the vast majority who are not needy—public programs are not the best or only expression of the public interest in economic security. Government should give them incentives to enhance their own economic security, without paying the freight charges to send their money round-trip to Washington.

There is a framework, then, for liberals and conservatives to argue about Good Government rather than Big Government. Start by rejecting all magic beans theories of public finance, acknowledging that the welfare state's expenditures and revenues cannot be massively, permanently out of balance. Both sides could accept the corresponding idea that the welfare state, like any utility-maximizing enterprise, has finite resources at its disposal. A well-designed and well-run welfare state will direct them to where they can do the most good and assist those people who need help most urgently. The corollary principle is that the welfare state should stop using its resources in ways that are not the best and highest, and reduce, reshape, or eliminate programs to help people who don't need it.

This framework leaves liberals and conservatives ample room to disagree about what it means for the welfare state to have sufficient resources, and to put them to their best use. What it does not allow them to pursue are plans to make America either a Scandinavian social democracy or a night-watchman state. Never, in our lifetimes, will America's welfare state become either a multiple or a fraction of what it is now. Such romantic visions of political transformation are not what the exigencies of domestic policy-making call for. In 1986, when the baby boomers' retirement was still merely a cloud on the horizon, James Q. Wilson wrote that 

we are in a painful period, when what people want from government vastly exceeds what they are willing to pay for. Politicians encouraged people to take this view and now must find a way to persuade them to take a different view…. It will be a life's work for some dedicated, patient, and determined people willing to stay the course.

* * *

For Correspondence on this essay, click here.