A Note on Budget Deficits

by Don Boudreaux on February 8, 2006

in Politics

Assume that the U.S. government more or less
reflects the informed will of the American people. (I believe this assumption to be unrealistic,
but I’ll put aside my skepticism for the moment.)

Being well-enough informed, American voters should know that
the true level of taxation is the amount of government expenditures – soon to
be about $2.77 trillion annually – and not the amount of tax revenue the
government collects.

So why are we concerned about Uncle Sam’s budget
deficit? If voters truly are well-enough
informed so that we can regard the expenditure side of the budget to reflect
not only the “will of the people” but “the informed will of the people,” why
not regard the revenue side (and the resulting budget deficit) with equal
equanimity? The entire budget –
expenditures, current tax revenues, and any need to finance a resulting deficit
with debt instruments – can be thought to be the product of sound and
acceptable, if not perfect, choice of the people.

If the entire package is what well-enough-informed voters
really want, why lament any part of it?

But if we recognize that the budget deficit is the
consequence of political dysfunction – voters either ignorantly or selfishly
(and dangerously for the country over the long run) loading an unjustified tax
burden on future taxpayers, why not regard the entire budget to be afflicted
with such dysfunction?


I imagine myself asking the above questions to, say, Paul
Krugman. Krugman (like many others)
rightfully criticizes today’s gargantuan government budget deficits. But he also (like many others) typically
assumes that most government programs and projects are justified – or at least
that these are programs that We the People in our wisdom have asked government
to deliver.

He can’t have it both ways. If government spending is justified because reasonably informed citizens
choose its content and level, then these same reasonably informed citizens must
also be assumed to choose the budget deficit. If, however, the budget deficit is the result not of informed citizen
choice but instead of dysfunctional political institutions, then these same
dysfunctional institutions must be assumed to operate on the expenditure side
of the budget.

Of course, it’s possible to say that citizen-voters are
sufficiently informed – and that politicians respond to these citizens’
preferences – but that these citizens also have a dangerously short
time-horizon, or that they callously disregard their children’s and
grandchildren’s futures. The budget
deficit, then, is a result of these regrettable features of voter preferences.

But if this is true, then why respect the expenditure side
of the budget? Won’t it, too, reflect
current voters’ dangerously short time-horizons?

Just asking.

Be Sociable, Share!



58 comments    Share Share    Print    Email


quadrupole February 8, 2006 at 1:21 am


I agree that it's government spending, by virtue of consuming *resources* that might otherwise be put to productive use, that truely taxes the economy. So the question becomes, what is the cost of different methods of *raising* that money. Right now it's about 4.5% (on ten year t-bills) if borrowed via deficit spending. Since we borrow about 3% of GDP currently, that puts our long cost at 0.135% of GDP (this year). So any tax policy that you propose that would slow GDP growth by more than 0.135% is a less efficient way of raising the money being spent than borrowing it.

One can even begin to mount an argument for larger budget deficits this way. Consider for example eliminating the tax on capital gains. Even if this doubled the budget deficit to 6% of GDP, we'd still be better off if it led to a 0.27% increase in GDP growth (or more).

99 February 8, 2006 at 1:55 am

quadrupole: I think your analysis is flawed, because you're only considering a single year. In the real world, of course, public sector borrrowing repeats itself year after year, and compounding interest makes the orginal sum (principle) borrowed grow ever larger. Those little bitty fractions of a point accumulate and grow. If we were to raise taxes by, say, $350 billion (or whatever the exact amount is to erase the deficit) we'd no doubt reduce growth quite considerably for a year or two. But a ways down the road — say five years — the economy would likely have returned to its original growth path. Indeed, my instinct tells me that if a rigorous pay-as-you-go strategy were stricly observed over the long term, the economy would be growing even faster than otherwise (provided care was taken not to overtax the economy via the accumulation of fiscal surpluses). I mean, if your way of thinking were valid, why not abolish taxation entirely in favor of borrowing? Surely that $2.3 trillion tax cut would boost growth, no?

Martin Geddes February 8, 2006 at 2:18 am

Isn't this really some kind of inter-generational warfare? You're assuming a homogenous "will of the people", when in fact elderly people enjoying their last few years appear rather keen for youngsters to under-consume and over-pay for in order to support their lifestyle.

quadrupole February 8, 2006 at 4:10 am


You are presuming that economic growth also doesn't compound. If we spend $350 billion dollars raised by taxing capital gains not only do we loose all of the return *this* year in that money being invested in economic growth, but all the growth it (and it's returns) might have returned in subsequent years.

Think of it this way, you have a pool of wealth in your economy (your GDP plus your current accounts deficit). Each of these dollars has a cost to take. No matter which dollar you take, it's cost compounds into the future. So the question is, which dollars are the most expensive and which the least? I would find it likely that taking capital gains dollars is more expensive than issuing new government debt as a way to acquire the dollars spent by government. It's very likely that corporate income taxes are also more expensive than debt.

Capital gains in 2000 where ~644 billion, which at our current 15% rate would yield about ~$100 billion. I doubt our current capital gains are higher than 2000 ;) Corporate taxes in 2005 where $278.3 billion. So eliminating both taxes would result in an additional ~$380 billion in deficit. , or about 3% of GDP. If we funded the elimination of these taxes with debt it would increase this years carry by 0.135% of GDP this year, and this would compound forward. If this leads to increased GDP growth (this year, and each subsequent year) that offset, then it would be a net gain to our economy. So the question is, do you think eliminating capital gains and corporate income tax would cause the economy to grown more than 0.135% faster?

quadrupole February 8, 2006 at 4:11 am


The thing is, everything is spending side. The intergenerational warfare is a *spending* side intergenerational war. It's not that young people are taxed, it's that government spends a lot on old people.

mark adams February 8, 2006 at 5:39 am

quadrupole, deficit spending does not mean we can actually borrow resources from the future. When our governments spend now it means higher interest rates divert spending from the private sector to the public in the present – which manifests itself in the same way as a tax.

Christopher Meisenzahl February 8, 2006 at 8:18 am

Don, take a look at this:

Budget Wish Lists Come and Go, But 'Entitlements' Outweigh All


Aaron Krowne February 8, 2006 at 9:04 am

I think this is a good observation and good argument.

Its the kind of argument that forces people to recognize the fundamental flaws of the governmental system, and of the idea of a benevolent, all-powerful state able to fix society's problems.

If the budget is balanced one year, what prevents it from going into deficit the next? The same system is still in place: voters are essentially left to beg that politicians do not over-spend the money that has already been confiscated.

Further, if your party is in office today and spending on your pet projects, what do you think is going to happen when your party is no longer in office? Do you think they'll remain forever? If so, why? The current situation proves that even a party gaining power on a platform of "virtue" will shortly become breathtakingly corrupted, in proportion to the additional power gained.

Randy February 8, 2006 at 9:51 am


Well said. But may I point out that you are assuming that future generations will actually agree to pay the debt. It seems to me that a default is at least as likely as higher taxes.

Don Boudreaux February 8, 2006 at 9:59 am


Perhaps future generations will default — but this move would be catastrophic, for Uncle Sam will no doubt manage to monetize his outstanding debt. Future generations will thus suffer the awful, heavy tax of inflation.

John F. Opie February 8, 2006 at 10:20 am

Hi -

Not quite.

It's not so much government expenditures that add to the tax burden, it's the financing of them that does. You're tending to mix apples and oranges here: government investments are multi-year, whereas tax revenues are per definition single-year. The servicing and retirement of debt is single-year and is therefore the correct term to use.

So the "real" tax burden is government income plus interest on total net government debt.

What you really should be looking at is government cash flow in order to see what the actual costs of government are and how that government debt is a) retired and b) rolled over and c) sustainability of that debt.

If you were to look at government as a business, which in many ways it is (non-profit of course), then only by analyzing the cash flow can you avoid any sort of ugly surprises like, for instance, balloon retirement of debt (like a mortgage, for instance, or multiple-year war bonds all coming due in a single year).

This is left as an exercise to the reader: suffice to say that debt is vastly better sustainable than you might think.


nn February 8, 2006 at 11:35 am

When I was in grad school, a prof who had been prominent as a govt advisor said frankly during one happy hour: The people mostly know what they want but it used to be filtered through the "right experts" (he meant the Kennedy brain trust, the NYT, the usual suspects) who knew how to implement and limit their demands. But now that Yahoos have taken over you have to fight them with universalist arguments while saving exceptions for the times the left gets into power.

Not those words, but the meaning was plain as day.

Ivan Kirigin February 8, 2006 at 11:35 am

The problem is raising revenues in a way that are growth-neutral. Raising capital gains taxes has a very clear effect on investment.

Perhaps raising gasoline taxes will not have as big of an effect, as recent sustained increases have shown.

If I were starting a country (look for Ivanistan in 20 years), funding would come through inflation, with strong constitutional checks on level of spending and areas of spending. All tax rates are explicitly zero, but implicitly are a flat rate against every unit of currency. There is no separation between spending and taxing — all spending has an immediate taxing effect that can't be removed. Clearly, minimal government would be needed.

Does anyone know a number for the effective increase in the money supply given printing and interest rates from the Fed? I wonder how close it is to $2.7T yearly.

quadrupole February 8, 2006 at 11:41 am

Mark Adams,

History does not bear out that increased government borrowing raises interest rates through crowding out, at least in the 3-5% of GDP range.

What really crowds out is government spending, independent of source, because it diverts funds from more productive uses.

Once you get beyond defense, contract enforcement, and law enforcement, what federal function exactly has a positive return on investment again? I would maintain that most federal government spending is a pure consumption, bearing no economic return. This is not to say that one cannot rationally argue that it is appropriate to trade growth for some other perceived good, but always keep in mind that government spending generally reduces growth.

Randy February 8, 2006 at 11:46 am


I agree that in the absence of sufficient tax revenue to maintain current levels of spending and payments on the debt, the government would likely monetize the debt.

But, assuming that there is an objective limit to revenue from taxes (Laffer Curve), then we are going to be in that situation eventually anyway. So why not apply a subjective limit (voters) now? There is a price to be paid for our overspending. We should pay it now.

Tim Lundeen February 8, 2006 at 5:44 pm

So there are two key factors to taxation: the level of government spending, and the structure of the taxation that supports it. Some tax structures have a significant negative impact on growth.

When people call for "repealing the Bush tax cuts", this is really an argument about the tax structure that is used to support our current level of spending. People who want to change the structure have an obligation to show that it will not reduce growth rates. It seems quite likely that increasing capital gains and rates on dividends will reduce growth rates. I suspect that increasing marginal personal income rates is also going to reduce growth, but the case here is not as strong.

quadrupole February 8, 2006 at 6:54 pm


I agree. But it's important to remember that borrowing is just another form of taxation of the economy.

It is also important to remember that growth need not be the only factor in the decision. I personally have a VERY strong bias in favor of growth. But it is legitimate to advocate (with open eyes) trading some growth for other considerations.

99 February 8, 2006 at 9:56 pm

****"I would find it likely that taking capital gains dollars is more expensive than issuing new government debt as a way to acquire the dollars spent by government."****

quadrupole: You'll get no argument from me here. My point wasn't focusing on the composition of a tax increase, but simply on the tax increase itself. I believe that deficit reduction brought about by a badly designed tax increase (like a badly designed tax code in general) will often not stand up to cost-benefit analysis. But what if deficit reduction is brought about by a tax increase that is designed in as benign a fashion as possible (say, increases in consumption or energy taxes)? I can't see any away around the logic that eliminating public sector borrowing, even if accomplished only on the revenue side, will generate net returns to wealth, growth and prosperity over the long term. Likely this is simply because, whatever short term damage a largeish tax increase would cause, eliminating government borrowing by definition increases national savings, and this should in turn tend to increase productivity.

Again, if I'm incorrect I don't see any reason why it wouldn't make sense to abolish taxation entirely in favor of borrowing.

Roy Stogner February 8, 2006 at 11:03 pm

Why it wouldn't make sense to abolish taxation entirely in favor of borrowing: because lenders make loans on the expectation that they will be paid back. If the planned means of payback is "I'll get a bigger loan from someone else to cover it, and a bigger loan from a third party to cover that, and…", then no sane lender will hand over money in the first place. Even the dullest of dot-com investors thought the companies they bought stock in would be paying dividends or buying back their own shares someday…

It seems to me that the most fundamental effect of government borrowing is pure supply-and-demand: you increase the demand for lenders, so you have to offer higher interest rates to compensate, and you have to (eventually, when your debt growth slows to equilibrium and you don't want to default) raise taxes to pay that interest. Essentially a government in debt takes money from people based on their income, and gives it out again based on their wealth. Combine this with cuts in inheritance taxes and capital gains taxes, and one might almost suspect that government policy was being made by people who had inherited more wealth than they earned…

Kevin February 8, 2006 at 11:55 pm

What's wrong with continuing to increase the public debt at a rate equal to growth in NOMINAL GDP? I figure, economy's doing pretty well now, the debt must be managable despite how scary the big number looks … if Nominal GDP grows at, say, 5.5%, why not run a fiscal deficit equal to 5.5% of the total debt?

If a poor guy carries an debt:income ratio of (say) 60%, surely a millionaire can afford to also?

I've never heard the issue posed like that, wondering if I'm missing some bad repercussion.

quadrupole February 9, 2006 at 1:51 am


Kevin points to your answer. As long as the deficit is lower than the growth in GDP, then our debt as a percentage of GDP declines, and we remain credit worthy. Since NOMINAL GDP growth is running around 6% or so, which means that eliminating capital gains and corporate income tax (and thus increasing the deficit to around 6% of GDP) wouldn't break the bank.

99 said:
"It seems to me that the most fundamental effect of government borrowing is pure supply-and-demand: you increase the demand for lenders, so you have to offer higher interest rates to compensate, and you have to (eventually, when your debt growth slows to equilibrium and you don't want to default) raise taxes to pay that interest."

That's all well and good… but when exactly have we seen rising deficits lead to increased interest rates? We've had many instances of high decifits now to look at, and yet we still get low interest rates in spite of them. I'm not quite sure what's going on there, but clearly higher deficits do not lead to higher interest rates historically. Unless you can describe why NOW is different from THEN, I don't see any reason to believe that higher deficits would lead to higher rates now.

spencer February 9, 2006 at 9:14 am

quadrupole — what about real interest rates? They have been much higher since 1980 when we started running much larger secular vs cyclical deficits. The historic cyclical increases in govt borrowing did not lead to increased rates because they were always accompanied by larger drops in private borrowing. But that type of cyclical deficits is really not what we are discussing now. Of course in this cycle we are still not seeing the normal increase in corporate borrowing.

quadrupole February 9, 2006 at 7:02 pm


Hmm… really…

I would be the first to point out that real interest rates are very hard to compute, because using the naive approach of looking at a particular years nominal interest rate minus that year's inflation rate doesn't really paint the whole picture. What moves nominal interest rates is the expectation of inflation. I would maintain that the higher real interest rates of the 1980s where primiarly a hang over from the high inflation expectations of the 1970s. If you look at the current situation compared to the pre-hyper inflation era of the 1970s you see that we have quite favorable 'naive' real interest rates:

In 2005, according to the Fed:
( http://www.federalreserve.gov/releases/h15/data/Annual_Dec_/H15_MORTG_NA.txt )
, nominal mortgage interest ran 5.86%, and inflation:
( http://inflationdata.com/inflation/Inflation_Rate/HistoricalInflation.aspx?dsInflation_currentPage=0 )
ran 3.39%, for a real mortgage interest rate of 2.47%. In 2005 the deficit was 2.6% of GDP.

Let's look at 1972 for comparison. Mortgage rates where 7.38%, inflation was 3.27%, for a real interest rate of 4.11%, clearly higher. In 1972 the deficit was 2.0% of GDP.

Having perused the historical data, I would say that the best way to acheive low real rates is either to suddently and unexpectedly hyperinflate (which I don't favor, as it works out badly in the long term) or to maintain low inflation over the long term (our current regime). I honestly don't think that government borrowing does much to move interest rates *unless* it begins to strongly increase the overall government debt as a percentage of GDP. The solution is growth, and thus in my mind anyone who REALLY cares about the long terms size of the debt and the deficit would favor borrowing to enable tax cuts that stimulate stronger growth than the deficit suppresses.

goof February 9, 2006 at 8:27 pm

Long term interest rates are obviously tied to deficits…What happens when the gov't runs a deficit? It issues more long term bonds to finance their expenditures and increasing the supply, decreasing the price and driving up the yield.

quadrupole February 9, 2006 at 8:46 pm


Then please riddle me the current 10 year bonds at 4.29%:


against an inflation rate of 3.39% for a real yield of 0.9%. Compare this to 2000 with a 10 year tbill rate of 6.03% and an inflation rate of 3.38% for a real yield of 2.65%. In 2000 we were running a 2.4% of GDP surplus, in 2005 we ran a 2.6% of GDP deficit.

Listen, I can follow the theory that deficits raise interest rates. The problem is that the empirical data doesn't bear that out. At all. I was brought up a physicist, when your data contradicts your theory, you go back and rexamine your theory. Deficits raising interest rates isn't born out by the data.

99 February 9, 2006 at 9:54 pm

***Kevin points to your answer. As long as the deficit is lower than the growth in GDP, then our debt as a percentage of GDP declines, and we remain credit worthy.***

I'm in agreement with this, although this implies a limit to your "borrowing is better than raising taxes argument". The government currently owes a net external debt of about $4.8 trillion. If the government adds to this an additional, say, $375 billion, it's increasing the size of its debt by about about 7.8 % in nominal terms, or perhaps 5% in real terms. The economy is growing pretty briskly, but not quite so briskly (in most years at least) as 5%. I suspect we would both agree that reducing this growth in the debt to where it is expanding no more rapidly than GDP as a whole would be a Good Thing. I also suspect we would both agree that the best way to accomplish this would be for the government to spend less money. But if such a course of action were not possible, I would argue that we would still derive some net benefit over the long term from accomplishing this goal via raising taxes. This is where I'm not sure you agree.

(Implicit in my argument, of course, and I'll state it explicitly here, is avoidance of growth in the government's share of GDP irrepsective of how government acquires its dollars; obviously if the government eliminated its need to borrow only via tax increases, and increased its take of national output while doing so, we would very definitely have on our hands a Bad Thing).

goof February 9, 2006 at 11:18 pm

ill grant that you may be right and the theory may be wrong…however, i think that anything conclusive requires a more sophisticated estimation and detection technique than quoting two samples and saying there is no correlation.

quadrupole February 10, 2006 at 5:34 am


I think we both agree that there is a wall beyond which deficits become a problem.

I don't think that wall is *quite* where we are increasing debt as percentage of GDP (since we hav increased from debt being 33% of GDP to 37.4% of GDP in recent years with no adverse effects). We've also had debt to GDP as high as 49.4% (in 1993) with little in the way of adverse effects. But I agree that there is a wall out there, and I feel that the wall is when our debt to GDP ratio grows to fast, or grows to large (I have no idea what 'to fast' and 'to large' may be).

Of course all of this would be moot if we could just keep spending to 17% of GDP :)

quadrupole February 10, 2006 at 5:39 am


Agreed, more than two data points are needed. I am just to lazy to compute them right now :)

If you look over the data for the 70s,80,90s, and, 00 one gets the sense that inflation rate expectations are sticky, and that they have a LOT to do with the 'naive' real interest rates. I suspect this is a much larger effect than anything related to deficit spending.

Previous post:

Next post: