“[W]e’ve got to reduce spending before we can reduce taxes. Well, if you’ve got a kid that’s extravagant, you can lecture him all you want to about his extravagance. Or you can cut his allowance and achieve the same end much quicker. But Government has never reduced. Government does not tax to get the money it needs. Government always needs the money it gets.” ~ Ronald Reagan
Hence was born the idea of “starve the beast,” a conservative core belief if ever there was one. But is it true? As explained in a recent column by Andrew Ferguson, economist (and libertarian) William Niskanen didn’t think so.

Beginning in 2002, Niskanen published a series of papers and op-eds about tax cuts and spending increases that turned conventional conservative wisdom on its head….If we wanted a smaller government, he said, we would have to raise taxes….Niskanen, looking over 25 years of budget data, noticed something about STB [“Starve The Beast” ~ ed.]: It didn’t work. In fact, attempts to starve the beast by tax cuts seemed to lead to increased federal spending.
Niskanen looked at both spending and taxes as a percentage of GDP. On average, he found, if federal revenues declined by 1 percent, federal spending increased by 0.15 percent. When revenues rose, on the other hand, relative spending decreased. A further study in 2009 by another Cato economist, Michael New, came to the same conclusion after the gluttonous administration of George W. Bush. Under Bush and his mostly Republican Congress, new benefits like subsidized Medicare drugs and increased federal education spending followed on the heels of large tax cuts.
Niskanen’s explanation for the failure of STB was straightforward, a conjecture based on standard economics: When you cut the price of something, demand for it will increase. Lowering taxes without lowering benefits meant that taxpayers were getting the benefits at a discount. The government made up the true cost with borrowed dollars that future taxpayers would have to repay. There was a big difference, Niskanen said, between a kid on an allowance and the federal government: The government has a credit card with no debt limit. {emphasis added}

That last–the ability of government to write checks on credit–was overlooked by STB advocates.

[E]arly advocates of STB had counted on something that never materialized. They had assumed that as the debt piled up to finance annual budget deficits caused by free-flowing benefits, public outrage would force politicians to restrain spending without raising taxes. Yet we’ve had the deficits and the borrowing, in amounts that would have left Friedman and Reagan agog; what’s been missing is the outrage.

People aren’t outraged because they don’t feel the immediate pain of increasing government because the money for government expansion is either borrowed or paid for by increasingly fewer individuals. So around 50% of the population feels no pain (they don’t pay income taxes) while a majority of the rest pays relatively minimal amounts. And a lot of that pain is left to future generations. This aligns with Niskanen’s reasoning for why higher tax rates lead to lower spending:

“Demand by current voters for federal spending,” he explained, “declines with the amount of this spending that is financed by current taxes.” When you make them pay for government benefits out of their own pockets, in other words, voters will want fewer of them. The journalist Jonathan Rauch put Niskanen’s point more pithily: “Voters will not shrink Big Government until they feel the pinch of its true cost.”

Yet, as I mentioned, not everyone shares the tax burden evenly in our progressive income tax system. So, perhaps a flat tax would prove or disprove Niskanen’s theory, but it’s doubtful that will happen any time soon.
Ferguson’s article brought some critiques of Niskanen’s ideas. Noah Glyn offers up another reason for why government spending decreases when tax revenue increases:

[It’s] the business cycle. As the economy grows, people earn more so they pay more in taxes; conversely, when the economy enters a recession, government revenue plummets. During recessions, however, the public relies on increased government spending, in the form of Medicaid, food stamps, and other transfer payments. (This can go the other way, too: Some state and local governments have used economic growth to justify increasing promises to government employees’ pension plans, but those costs typically come much further down the line.)

This is buttressed by Ramesh Ponnuru’s important, technical point and “thought experiment”:

Let’s say we still had the Clinton-era tax rates and a (smaller but still quite large) long-term debt problem. Wouldn’t we be debating an increase in tax rates to a higher level than we are now? That seems to me pretty likely. The baseline from which we’re negotiating would be higher, perceptions of what’s tolerable would be higher, expectations of tax rates would be higher. On the Niskanen theory there would be a countervailing effect: In the interim the tax cuts caused spending to be higher and thus moved the spending baseline higher. But Niskanen didn’t find that a dollar of tax cuts were associated with a dollar of spending increases; he found that a 1 percent reduction in revenue over GDP was associated with a 0.15 percent increase in spending over GDP. So the countervailing effect would be smaller.

Jonah Goldberg adds:

I always liked Niskanen’s argument, even if I didn’t quite find it persuasive. One thing that always bugged me about it which, to my surprise, Ferguson doesn’t mention, is the implicit assumption that Americans behave like rational economic actors with regard to what they get from government….The American species of homo economicus has been paying hundreds of billions to get rid of poverty for decades, what do we have to show for it? Poverty rate in 1975: 26 percent. Poverty rate in 2010: 26 percent. What a great return on the investment. Federal spending on education? Ahem…For reasons, good and bad, voters don’t treat tax dollars the way they do their own dollars. They don’t demand quality. They don’t demand accountability. They don’t push for efficiency. Many people think the government should spend money as if it comes from someplace other than the wallets of citizens and that what we get for it should be graded on some spiritual, emotional, philanthropic or metaphysical curve. How we spend for X so often seems to matter more than how much X is actually delivered.

Yet, as Patrick Brennan argues, re-stating Niskanen’s implicit premise, the missing demand for government quality is because so many have so little stake in the game.

People might be a lot more likely to start caring about where their tax dollars go (whether the ends are efficient and whether the money comes back to them) when those taxes are really substantial, broad-based, and they actually have to pay them.

Brennan also compares U.S. expectations for government services to that of Europeans:

If you live in a society where, as Jonah pointed out Arthur Brooks has argued, the state is considered the main conduit for meeting societal needs and caring for the poor and vulnerable, you’ll care more about how well government works and whether it can care competently for you, and that’s a cultural matter. But it’s also important to homo economicus, because Leviathan has taken most of his paycheck, and he now has to hope, and should ensure, that government will provide for society at large, the poor and vulnerable, and even him at times, and do so as efficiently and competently as possible.
There are obviously other explanations for these differences: Charlie Cooke has lamented to me on many an occasion that in Britain, the conversation about almost all government policies ends up being debates over efficacy of programs, not whether the programs should exist in the first place. Leaving aside the financial constraints Britain and elsewhere are now experiencing, if you don’t have a constitution with enumerated federal powers, a truly conservative and independently minded political movement, etc., you’re going to spend more time on making government work, not on making it smaller, and that’s for other reasons than I’ve just proposed.*

Regarding the last, many conservatives (well, at least me) believe that a smaller government is one that is easier to make workable!
==========================
* Brennan expounded on the point later in the post: “It’s difficult to assess my thesis inasmuch as big government and the cultures that give rise to it have other negative effects on efficiency, so it’s possible citizens subject to a huge government and a regressive tax code get a more efficient government than they would if they didn’t have higher expectations than free-riding Americans, but still not a very efficient one. It’s been suggested, in fact, that it’s highly efficient yet regressive taxation (like light capital taxation, competitive corporate-tax rates, consumption taxes, etc.) that’s allowed places such as France and Scandinavia to have functional economies despite the burdens of absurdly large governments; perhaps it’s also the relative efficiency and usefulness of their government spending programs, and not just their tax system, that’s allowed them to manage as well. Thus again, economic preferences force the hand of citizens and politicians in a completely government-dominated society but not in one like America.”