Receptive as I usually am to arguments from Milton Friedman, "starve the beast" has always puzzled me. All governments would like to cut taxes. You win elections that way. What stops them is a mixture of two things: They have spending plans they want to finance, and they recognize limits to how far they can let the budget deficit expand. Concern for the deficit is crucial. If the government thinks that the deficit is a nonissue, lack of revenue would not check its spending ambitions. It could spend more and just keep borrowing. For "starve the beast" to work, governments have to worry that the deficit will get too big.
Here is the puzzle: Advocates of "starve the beast" tend to be the very politicians and economists who do not care much about the deficit. They want to cut taxes even when the government is already borrowing heavily, partly (they say) to put further downward pressure on spending. The question is, what downward pressure? What is wrong, in their view, with simply running a bigger deficit? Nothing, really. Given the choice between a) raising taxes and curbing the deficit, and b) cutting taxes and watching the deficit grow, they would always choose b.
It makes no difference to the beast whether you feed it extra taxes or shovel in extra borrowing: It keeps on eating, regardless. If you dismiss concern about the budget deficit, where is the constraint that provides the fiscal discipline to cut spending? There isn't one. On the other hand, if you do care about the deficit, you would not be advocating a tax cut in the first place, unless you were calling for spending reductions at the same time. The whole "starve the beast" idea seems self-refuting.
Well, not necessarily. One way to rescue it, from a logical point of view, is to suppose that people are playing games. The Republicans who are keen on "starve the beast" may be thinking: "We don't care much about the budget deficit, but those foolish Democrats do.
Let's use that. If we can get taxes down, and the deficit goes up, Democrats will grow more amenable to spending cuts. That way we will end up with lower taxes and lower spending as well."
For the country as a whole, you could argue, this would be a good result, albeit a paradoxical one. The Republicans' fiscal recklessness turns the superior fiscal prudence of the Democrats against the Democratic inclination to spend and tax too much. Democratic priorities on taxes and spending are undone by the party's own virtue on the deficit. An old-fashioned fiscal conservative such as myself -- who prefers low spending, low taxes, and a balanced budget -- might look at it this way: It hardly seems fair, but it just might work.
Federal spending declined as a share of national income during President Clinton's two terms in the 1990s. Fiscal responsibility -- the need to curb the budget deficit -- was a consistent theme. The deficits that Clinton inherited, which put the squeeze on spending, were bequeathed by the earlier large tax cuts (and increases in defense spending) of the Reagan administration. So perhaps the strategy worked after all.
Several statistical studies of "starve the beast" have gathered more systematic evidence than that, but unfortunately they have not succeeded in firmly settling the issue.
A new study by Richard Vedder and Jonathan Leirer for the congressional Joint Economic Committee, for instance, found that over the past 60 years every dollar raised in extra taxes has given rise to more than a dollar in extra spending. "The findings," say the authors, "cast grave doubt on the efficacy of raising taxes as a means of eliminating fiscal imbalances."
Although this study looks at the issue from the other side (a tax increase rather than a tax cut), this result is consistent with "starve the beast" because it suggests that new taxes don't cut total government spending; and it is a fairly representative finding. The phenomenon is also relatively modern, the authors argue. The federal propensity to spend new tax revenues, rather than use them for deficit reduction, has trended upward ever since the days of George Washington. "At one time, new taxes were associated with very significant deficit reduction, but not in recent decades."
Another new study, however, points the other way, and it deserves to be taken seriously because it is an unusually thorough analysis. Christina Romer and David Romer, both of the University of California (Berkeley), find "no support for the hypothesis that tax cuts restrain government spending" and conclude that "the main effect of legislated tax changes on the overall government budget is to induce subsequent tax changes in the other direction."
The post-Reagan years also illustrate this second point. The Tax Equity and Fiscal Responsibility Act of 1982 immediately clawed back some of the cuts of 1981. Further tax increases followed in 1984, 1987, and 1990 (Papa Bush's "no new taxes" notwithstanding). Sounding like a New Democrat, or an old-school Republican, Ronald Reagan explained in 1982 that the choice was to "reduce deficits and interest rates by raising revenue from those who are not now paying their fair share," or to "accept bigger budget deficits, higher interest rates, and higher unemployment." It is true that the deficit was still there in 1992, and was a main reason for spending restraint during the Clinton administration -- but on closer examination this is not the clear-cut case of "starve the beast" that you might suppose.
What about the tax cuts of 2001 and 2003? It would be hard to argue that spending restraint has been among the current Bush administration's subsequent priorities: "Big-government conservatism" about sums things up. Further pressure to expand the budget deficit is coming, thanks to demographic and other forces; but whether or when that problem will be addressed, and if so by spending cuts, tax increases, or some of both, remains to be seen. The Bush tax cuts are unfinished business.
In rejecting "starve the beast," the Romers make an important point about what might be wrong with the studies that say it works. In looking at changes in taxes and spending over the past 60 years, they find many instances of tax changes that were motivated by new spending policies introduced at the same time -- that is, packages of tax and spending cuts, and packages of tax and spending hikes. There are enough of these in the sample, they say, to skew the results and give a false indication that tax cuts encourage spending cuts.
"Starve the beast" exponents are not demanding packages of lower taxes and lower spending. They are saying that lower taxes will sooner or later wear spending down anyway. When you look at those cases -- instances where taxes have been cut independently, with no connection to new spending plans -- spending does not fall, say the Romers. In fact, it rises a bit. "Starve the beast" does not work.
I hope they are right. The idea that a kind of political extortion is needed to contain the growth of government may possibly be correct -- but it is certainly unappealing. The case for low taxes can be made perfectly well on the merits: Arguments one (pro-growth) and two (pro-liberty), as mentioned at the start, ought to be enough. Then, if you can convince people that persistent large budget deficits are bad for the economy (which they are), the case for limited government is made as well.
The worst thing about "starve the beast" is the idea that a straightforward argument for low taxes and spending cannot be pressed successfully. You have to cut taxes, which the voters will like, and let them think they can have high spending, too. Later, if all goes according to plan, they will see they were mistaken. It is a strategy based not only on outwitting the Democrats, but on outwitting the electorate as well. It would be a pity if it worked.