Friday, February 24, 2006

Starving the Beast, Updated

Out-of-control spending is a hot topic of conversation in the blogosphere, especially among those who are disappointed in President Bush's failure to curb the federal government's appetite. Bill Niskanen and Peter Van Doren, former colleagues of mine at Cato Institute, published a paper a few years ago (which no longer seems to be available on the web), in which they said this:
For nearly three decades, many conservatives and libertarians have argued that reducing federal tax rates, in addition to increasing long-term economic growth, would reduce the growth of federal spending by "starving the beast." This position has recently been endorsed, for example, by Nobel laureates Milton Friedman and Gary Becker in separate Wall Street Journal columns in 2003.
It seems to me that the notion of starving the beast is really an outgrowth of an older, simpler notion that could have been called "strangle the beast." The notion was (and still is, in some quarters) that the intrusive civilian agencies of the federal government, which have grown rampantly since the 1930s, ought to be slashed, if not abolished. There's no need for fancy tricks like cutting taxes first, just grab the beast by the budget and choke it. There's more than money at stake, of course -- there's liberty and economic growth. (I have shown here the extent to which the beast of government has strangled economic growth.)

Anyway, Niskanen and Van Doren argue that the "starve the beast" strategy has failed, which is true, but I have serious reservations about their analysis. Their figure of merit is spending as a share of GDP. But it's the absolute, real size of the beast's budget that matters. Bigger is bigger -- and bigger agencies can cause more mischief than smaller ones. So, my figure of merit is real growth in nondefense spending.

What about defense spending, which Niskanen and Van Doren lump with nondefense spending in their analysis? Real nondefense spending has risen almost without interruption since 1932, with the only significant exception coming in 1940-5, when World War II cured the Depression and drastically changed our spending priorities. Real defense spending, on the other hand, has risen and fallen several times since 1932, in response to exogenous factors, namely, the need to fight hot wars and win a cold one. Niskanen and Van Doren glibly dismissed the essentially exogenous nature of defense spending by saying
that the prospect for a major war has been substantially higher under a unified government. American participation in every war in which the ground combat lasted more than a few days -- from the war of 1812 to the current war in Iraq -- was initiated by a unified government. One general reason is that each party in a divided government has the opportunity to block the most divisive measures proposed by the other party.
First, defense outlays increased markedly through most of Reagan's presidency, even though a major war was never imminent. The buildup served a strategy that led to the eventual downfall of the USSR. Reagan, by the way, lived with divided government throughout his presidency. Second, wars are usually (not always, but usually) broadly popular when they begin. Can you imagine a Republican Congress trying to block a declaration of war after the Japanese had bombed Pearl Harbor? Can you imagine a Democrat Congress trying to block Bush II's foray into Afghanistan after 9/11? For that matter, can you imagine a Democrat-controlled Congress blocking Bush I's Gulf War Resolution? Well, Congress was then in the hands of Democrats and Congress nevertheless authorized the Gulf War. Niskanen and Van Doren seem to dismiss this counter-example because the ground war lasted only 100 hours. But we fielded a massive force for the Gulf War (it was no Grenada), and we certainly didn't expect the ground war to end so quickly.

As I was saying, domestic spending is the beast to be strangled. (I'm putting aside here the "sacred beasts" that are financed by transfer payments: Social Security, Medicare, etc.) How has the domestic beast fared over past 70-odd years? Quite well, thank you. It fared best from 1933 through 1969, but it hasn't done badly since 1969.

The beast -- a creature of the New Deal -- grew four-fold from 1932 through 1940. Preparations for war, and war itself, brought an end to the Great Depression and stifled nondefense spending: It actually dropped by more than 50 percent (in real terms) from 1940 through 1945.

After World War II, Truman and the Democrats in control of Congress were still under the spell of their Depression-inspired belief in the efficacy of big government and counter-cyclical fiscal policy. The post-war recession helped their cause, because most Americans feared a return of the Great Depression, which was still a vivid memory. Real nondefense spending increased by 180 percent during the Truman years.

The excesses of the Truman years caused a backlash against "big government" that the popular Eisenhower was able to exploit, to a degree, in spite of divided government. Real domestic spending went up by only 9 percent during Ike's presidency.

The last burst of the New Deal came in the emotional aftermath of Kennedy's assassination and Lyndon Johnson's subsequent landslide victory in the election of 1964. Real nondefense spending in the Kennedy-Johnson years rose by 56 percent. The decades-long war over domestic spending really ended with the enactment of LBJ's Great Society. The big spenders won that war -- big time. Real nondenfense spending grew at an annual rate of 5.9 percent from 1932 through 1969.

Real nondefense spending has continued to grow since 1969, but at the lower rate of 2.5 percent per annum. What has changed is that nondefense spending has grown more steadily than it did from 1932 to 1969. Each administration since 1969 (aided and abetted by Congress, of course) has increased nondefense spending by following an implicit formula. That formula has two parts. First, there is the steady increase that is required to feed the beast that came to maturity with the Great Society. Second, there is countercyclical spending which is triggered by recessions and unemployment. As a result, there is a very strong -- almost perfect -- relationship between real nondefense spending and the unemployment rate for the years 1969 through 2005. Using a linear regression with six pairs of observations, one pair for each administration, I find that the percentage change in real nondefense spending is a linear function of the change in the unemployment rate. Specifically:
S = 1.0277 + 0.11346U

where S = real nondefense spending at end of a presidency/real nondefense spending at beginning of a presidency

U = unemployment rate at end of a presidency/unemployment rate at beginning of a presidency.

The adjusted R-squared for the regression is .979. The t-stats are 112.17 for the constant term and 15.26 for U.
What about divided government, of which Niskanen and Van Doren are so fond? Divided government certainly hampered the ability of Republican administrations (Nixon-Ford, Reagan, Bush I, and Bush II) to strangle the beast, had they wanted to. But it's not clear that they wanted to very badly. Nixon was, above all, a pragmatist. Moreover, he was preoccupied by foreign affairs (including the extrication of the U.S. from Vietnam), and then by Watergate. Ford was only a caretaker president, and too "nice" into the bargain. Reagan talked a good game, but he had to swallow increases in nondefense spending as the price of his defense buildup. Bush I simply lacked the will and the power to strangle the beast. Bush II may have had the power (at one time), but he spent it on support for his foreign policy and in an effort to buy votes for the GOP.

Bureaucratic politics (rather than party politics) is the key to the steady growth of nondefense spending. It's hard to strangle a domestic agency once it has been established. Most domestic agencies have vocal and influential constituencies, in Congress and amongst the populace. Then there are the presidential appointees who run the bureaucracies. Even Republican appointees usually come to feel "ownership" of the bureaucracies they're tapped to lead. Real nondefense spending therefore has risen steadily from the Great Society baseline, fluctuating slightly in countercyclical response to recessions and unemployment.

Having said all that, how do the presidents from Nixon through Bush II stack up? In spite of all the blather about Bush II's big-spending ways, there's not a dime's worth of difference among the post-Great Society administrations -- Democrat or Republican. Using the above regression equation, I estimated the expected growth of real nondefense spending for each administration. I then used that estimate to compute an actual-to-estimated ratio (how much nondefense spending actually rose divided by how much it "should" have risen, according to the equation).* The results:
Nixon-Ford -- 1.00
Carter -- 1.00
Reagan -- 1.00
Bush I -- 1.00
Clinton -- 1.01
Bush II -- 0.99
The lesson is clear: Tax cuts won't starve the beast -- Friedman, Becker, and other eminent economists to the contrary. But tax increases, on the other hand, would only stimulate the beast's appetite. The best way to cut spending is . . . to cut spending.

In any event, the truly vicious beast isn't federal nondefense spending, it's state and local spending. Spending by state and local governments in the United States is five times as large as the federal government's nondefense spending. Real spending by state and local governments increased by a multiple of 11 from 1945 to 2005. The population of the United States merely doubled in that same period. Thus the average American's real tax bill for state and local government is more than five times larger today than it was in 1945.

It's evident that not enough of the loot has been spent on courts and police. No, our modern, "relevant" state and local governments have seen fit to waste our money on such things as free bike trails for yuppies, free concerts that mainly attract people who can afford to pay for their own entertainment, all kinds of health services, housing subsidies, support for the "arts," public access channels on cable TV, grandiose edifices in which our state and local governments hatch and oversee their grandiose schemes, and much, much, more.

Then there are those public schools . . .

The good news about state and local spending is that its real rate of growth has dropped in the past few years. The bad news is that the slowdown coincided with a recession and period of slow economic recovery. The good news is that state and local spending is a beast with thousands of necks, and each of them can be throttled at the state and local level, given the will to do so.
* By the standards of 1969-2005, here's how earlier administrations stack up:
Hoover -- 0.68
Roosevelt -- 3.29 (through 1940)
Truman -- 2.30
Eisenhower -- 0.85
Kennedy-Johnson -- 1.43
There used to be a real difference between Republicans and Democrats. Now there isn't. That doesn't make Democrats any better, it just confirms my version of the old adage: The pursuit of power corrupts.