Arnold Kling  

A Basic Divide

The Time Capsule from 1983... Why Hillary's Harmless...

Greg Mankiw writes,

If you think it is the job of government to take from Peter to pay Paul, and if Peter can move around the globe, then you need international tax cooperation. Otherwise, some countries will become nations of Peters, leaving all the Pauls to fend for themselves. On the other hand, if you think that the main job of government is to facilitate voluntary exchange by protecting property rights, rather than re-slicing the economic pie as it sees fit, then tax competition is a good check against excessive interventionism. In other words, are you more worried about too little government or too much?

The question that I would ask Larry Summers (who is on the other side from me on this issue) would be this: how big a share of GDP would the Federal government have to control before you would worry about government being too big? I mean, unless you have a totally romantic notion of government as the embodiment of the will of the people (and I don't think Larry would make that defense), there must be some upper limit. Right now, the Federal government directly controls roughly 20 percent of GDP, and indirectly it controls more through regulation, mandates for state governments. etc. If that's not too much, then what is? 30 percent of GDP? 50 percent of GDP? 75 percent of GDP?

Health care, the fastest growing major sector of the economy, is also the sector that is increasingly being absorbed by the government. Extrapolating current trends, Medicare alone would be most of GDP in another generation or so. It boggles my mind how someone can not be worried about too much government.

Comments and Sharing

CATEGORIES: Political Economy

COMMENTS (18 to date)
mgroves writes:

But, as long as [insert party I belong to] is in power, things will be fine!

Lord writes:

If you want to compare us to the rest of the industrialized world, apparently quite a bit more.

brian writes:

Why should the answer be in the form of a percentage? That seems ridiculous to me. The most important thing from an economist's point of view should be efficiency, not what "percentage" is technically under government "control."

To see why, consider tax subsidies: if we eliminated all distortionary tax subsidies and exemptions, government control as a percentage of GDP would go up, but economic efficiency would surely be increased. And I expect that Arnold would think this is a good thing, because the government is less distortive--despite the fact that the percentage is larger.

What's important is not that percentage but the quality of policy. What's worse: 20% of the economy devoted to distortionary farming subsidies or 20% of the economy devoted to addressing public goods? The answer should be obvious, which should tell you that the measure of the optimal amount of government is not solely a function of its size, but (perhaps more importantly) of its quality.

Thus, Arnold's question has no specific answer. You have to look at the finer issues involved, which involves going beyond a one-word answer.

wintercow20 writes:

Ceteris paribus Arnold's question sure does have an answer.

That aside, efficiency arguments lead one down a dangerous path. If everyone in "society" decides that 5'4" Italian-Americans from Woodhaven, NY are a pox on society (even if only a little bit), and if I happen to object to this contention, don't efficiency arguments send me up the river?(at least the ones that seem to rule today - i.e. Kaldor Hicks, not Pareto)

A respected friend tells me repeatedly, anyway, that there is no constituency for efficiency outside of an economics classroom (in Washington).

Matt writes:

When Peter is worried about excess government he wants his progressive taxes reduced. When Paul is worried about too little government he wants progressive taxes raised. We do the balance.

Why do Peter and Paul think the mechanism is progressive taxes? Because it is all they have available to them to get at the market equilibrium of government. They both want to match the flow of government products to their relative wealth. They are cooperating, within their ability, to over come the quantum limit of one man per one vote.

The income tax was designed specifically as the best mechanism we have to get government into equilibrium. When a libertarian criticizes progressive taxes, ask him a better way to equilibriate government. He will say, compute the solution! But that is unlibertarian, is it a social planner result not a set of differential moves.

8 writes:

If we removed all distortionary tax subsidies and exemptions, the total tax rate should fall but government taxes as a percent of GDP would be neutral. If the government just closed loopholes, assuming higher efficiency means the new government spending is less distortionary than the existing subsidy, but where's the evidence for that?

If I take Mankiw's argument to the extreme, the best way for the U.S. to export its system and maintain global hegemony would be to wage a war of tax attrition. It could also set the world on a path to minarchy.

8 writes:

over come the quantum limit of one man per one vote.

Fundamentally that's the problem, not the tax rate.

aaron writes:

I think the obvious answer is we should grow the government up until the point where any additional growth has a negative impact on GDP.

When tax increases lead to GDP growth (because of high interest on debt and good choices in government spending), we should raise taxes. If tax cuts lead to increases in GDP (due to low interest and poor governemnt spending decisions)we should do that.

Any increase in GDP utimately will also lead to higher revenue for the government too. It's win-win (OK, sure. The current politician will be long dead by the time revenues excede what they would have been otherwise, but politicians are altruistic, aren't they?).

John Fast writes:

1. This is another post where the chapter of _The Machinery of Freedom_ with the thought experiment about living in housetrailers is relevant.

2. Lord, I agree with you and suggest two tweaks to Dr. Kling's question. First, we should consider the total share of all government spending, i.e. state and local, not just the federal government. Second, most other developed countries have "responsible-party" systems, which reduce government waste and stupidity, at least somewhat.

3. Wintercow20, I agree with you about the rights of minorities and individuals. That's one reason it's so important to start with a fair distribution of rights, and also why we should use Pareto efficiency, rather than Kaldor-Hicks efficiency, i.e. compensate people for any harms they suffer.

And there is always a constituency for anything, if it's packaged properly. People will go for economic efficiency if the benefits are explained clearly.

Lord writes:

Since there is little in the way of real economic growth in this country, healthcare may be the solution rather than the problem. Perhaps some alternative energy investments will deliver in the future, but until then, it may be all we have.

larry writes:


The risk to a democratic government with a progressive income tax is that when 45% of the electorate is comprised of taxpayers and 55% of tax consumers who pay no federal income tax, the 55% can outvote the 45% every time. They don't pay for tax increases, and presumably they get something from more spending. That creates a powerful constituency for more and more spending.

Matt writes:


"presumably they [55%]get something from more spending [ than the 45%]. "

The key word is presumably. How presumed are they?

At some point, government is so big that even a large batched of the presumes will switch. And if we can prove that, then we know we will have smooth government equilibrium over, at least, very long periods.

It is the inelasticity of substitution of government for other services that prevents the presumes from getting an earlier signal. That inelasticity is not a killer, it is a slower downer. Having no elasticity at all is the killer, we have some at least.

We need to make the system give earlier signals. Who knows best when government is too big? Who knows when it is too small? Find those people, and we can curve their elasticity appropriately.

Helder writes:

An answer:

GarGi Dixit writes:

Too much government is extreme danger. And it is real.
As Ayn Rand Out it, "We are fast approaching the stage of the ultimate inversion: the stage where the government is free to do anything it pleases, while the citizens may act only by permission; which is the stage of the darkest periods of human history, the stage of rule by brute force."

In UK you cannot slap your own child, as its against children's right. No matters children are going for drugs and prostitution.

The rising prices of food grains are visibly because of the ill policies of government, but government keep criticizing and blaming each other.
On the name of social medicare system Canadian's are being victimized by government.
Things are surely getting worse.
To hide out their ill policies, they create new illusions everyday and now in form of Global warming and related catastrophe they tries to illude and deviate human from aspiring for Individual freedom with dignified individual rights based on reason.

Snark writes:

I disagree with those who advocate for larger government on the basis of economic growth or prosperity. It looks good on paper, as do strong leadership, political stability, and national security. However, these are precisely what Nazi Germany experienced pre-WWII and what communist China is currently experiencing. I don't believe freedom and liberty are the antithesis of Pareto efficiency and a fair distribution of rights that requires government intervention to correct.

John Fast writes:

Snark, I'm not sure if your comment was a reply to mine. In case it wasn't clear, I believe reducing/replacing government as much as possible will benefit the public as a whole, and will give the most benefit to the poor, disadvantaged, and oppressed.

I also favor using Pareto efficiency as a way to enforce human rights, especially the rights of the oppressed.

Chuck writes:

Because markets are efficient, the less government interferes with markets, the more likely it is that poor decision makers will have little money.

Then, because good decision makers have control over more capital, our GDP will rise very very quickly, benefitting all.

It is undeniably logical that pro-growth policies would indicate a redistributionist tax program where the higher your income, the *less* money you owe the government, since your high income indicates your skill at investing your capital to societies benefit.

Jason writes:

There will always be people in need. I am not sure if the government is the most efficient way of helping people. Non-profits and Churches should play a larger roll in helping people. The government does play a role in protecting people and their property and we will always need some source tax to implement this. Some of the functions that the government is filling now should be filled by Churches and Non-Profits. I am not sure what the best way to implement this is but I would allow people donate portion of their income tax to a non-profit.

Comments for this entry have been closed
Return to top