February 13, 2003

silhouette3.JPG From the desk of Jane Galt:

Do Deficits Constrain Spending?

Long time comments resident Jason, a math major (faugh!) sends this letter:
I ran some correlations on the budget data in this comment thread of Delong's:

Summary:

1) The correlation between surpluses and non-mandatory non-defense spending
is weakly negative.

2) There is no correlation between revenue and non-mandatory non-defense
spending.

3) The correlation between interest outlays and non-mandatory non-defense
spending is weakly negative.

1) & 2) contradict the Friedman hypothesis on deficits constraining
spending, and 3) supports it. However, why the hell would interest outlays
be a good measurement?
Math majors. . . ;-)

Because you're not factoring in expectations.

It's unlikely that a budget deficit in any one year will constrain spending, for the same reason that, say, having a wedding doesn't mean that most people stop going out to dinner or downgrade their car. The availability of credit in the 20th century has improved the ability of people to smooth their consumption over their lifespan, so when they have a temporary downturn, they borrow and spend, while when they have a temporary increase, they save a lot of it to spread the consumption over their lives (which is why temporary tax cuts such as the Democrats are proposing don't work very well as stimulus. IF you believe in stimulus.)

The US government does the same thing. If we have an enormous expense one year, such as, say, a war, we don't cut domestic spending or raise taxes by the entire amount, because that would cause a shock to the economy, and more importantly, to the voters. Rather, we try to spread out the pain over many years so that the impact in any one year will be relatively small. Thus, you wouldn't expect to see much correlation between the current budget deficit and spending.

Where you would expect to see a correlation is between larger measures of the government's future ability to spend. Basically, the voters, like us, are broadly interested in smoothing consumption of government goods over their expected lives. Spend too much now and they worry about later. Both interest payments and interest rates are a good indication of constraints on future government spending -- interest rates, because they raise the cost of consumption today in terms of consumption tomorrow, and interest payments, because they lower the amount of expected consumption in the future. Think of it in terms of your own life. When interest rates are low, you're happy to borrow, because the cost of consumption now is relatively low in terms of the amount you'll have to sacrifice later to pay it off. When interest rates are high, on the other hand, each additional dollar you borrow means giving up a lot of future consumption, so you constrict spending now. Similarly, if your student loan payments are $100 a month for the next thirty years, you probably do a lot more spending than if they're five times that.

You're getting a correlation between surpluses and spending because, weakly, when spending goes down surpluses go up.

I'd also caution that it's awfully hard to do realistic regressions on stuff like this because the data is thin. We've only had a hundred budgets over the last century, after all. Statistically, a sample population of less than 50 is usually considered to be near-useless; too much noise, not enough signal. But if you go back farther than 50 years, the political climate has changed so much that it's very hard to be sure you're comparing apples to apples -- with the Federal budget at 3% of GDP, can we make reasonable comparisons between Calvin Coolidge's budgets and ours? How do you deal with WWII, when the war consumed over 50% of GDP? What about entitlements? Nonetheless, the data broadly tracks what we'd expect to find, and it's certainly an interesting experiment.

Posted by Jane Galt at February 13, 2003 08:50 AM | TrackBack | Technorati inbound links
Comments

I posted a clearer summary here.

Posted by: Jason McCullough on February 13, 2003 09:06 AM

Aren't you implicitly assuming deficits affect interest rates, too?

Posted by: Jason McCullough on February 13, 2003 09:48 AM

Hmm... the original argument was 'tax cuts constrain spending', which then became ‘deficits constrain spending’ , which upon closer inspection turns out to be ‘high interest rates and interest payments constrain spending’. Which, as Jason points out, kind of runs afoul of the whole ‘deficits don’t affect interest rates’ thing. But whatever.

It’s important to view this whole issue in the political context, which is that Republicans, instead of being the principled advocates of smaller government, have from Reagan onward become the ‘All Tax Cuts, No Sacrifice’ party. In order to justify this economically dishonest and incoherent philosophy, they shift back and forth between three rationales:

1. Tax cuts won’t cause deficits, because the will increase economic growth.

2. Actually, tax cuts won’t cause deficits because they force spending restraint.

3. Anyway, what’s so bad about deficits? They aren’t really that harmful anyway.

What makes me want to beat my head against the wall is how people who ought to know better and are normally quite sensible (like you Jane) defend this sort of ridiculousness.

Posted by: RC on February 13, 2003 10:57 AM

Defend what? He asked me a technical question. I tried to answer it. I haven't commented on whether or not interest rates are raised by deficits -- for the record, I think they are, and I also think that given the generally short window of budgetary expectations in the market, the effect of endogenous budget reductions under Clinton, rather than exogenous capital gains increases, are fairly trivial and certainly not what the Democrats like to claim in the interests of crediting Clinton for the economic boom. But that's neither here nor there -- Jason asked why the correlations might work out the way they do, and I gave my opinion, along with my opinion that it's very, very hard to do a meaningful regression on US budgetary behavior because the sample is always to small.

Posted by: Jane Galt on February 13, 2003 11:18 AM

if you look at late 90s/2000 budgeting, you can see that continued surpluses correlate strongly with massive increases in government spending (incluiding incresing entitlement benefits)

course once they're implemented the economy is back in the crapper... so yeah, program years don't correlate well with econ years, due to lag in implementation...

couse wouldn't this seem to show why spending increases are bad and cuts are good (i.e. you get a lag, and then get your result predicted)

as for current situation.. well you've got a slowdown and a war, so yeah you're going to have problems in the finance department... course without tax cuts we'd be in surplus next year and the dems would be screaming for free lexus for seniors, or something... so tax cuts are very likely reducing gov't growth...

Posted by: Libertarian Uber Alles on February 13, 2003 11:52 AM

Isn't there something inherently dishonest in this "increase deficits to constrain spending" strategy, whether it works or not?

If you think the government spends too much, then make the case and say what you want to cut. And none of this Reaganite "future unspecified cuts" business. Don't hand out the tax-cut goodies to your buddies today and expect future leaders to sort out the mess.
Didn't Bush say something in the SOTU about not passing on our problems to future generations?

Posted by: Bernard Yomtov on February 13, 2003 12:09 PM

"if you look at late 90s/2000 budgeting, you can see that continued surpluses correlate strongly with massive increases in government spending (incluiding incresing entitlement benefits)"

Two things to think about:

-Does the correlation hold if you adjust spending for inflation and population growth?

-Does the correlation hold if you look at a longer period?

If both of these are not true, it's unlikely the correlation is meaningful.

Posted by: Mark on February 13, 2003 12:29 PM

I think it is pretty clear that neither party really cares about deficits except as a weapon to use against their opponents. The Republicans want tax cuts and to stay in power, and are willing to spend enough to blunt the Democrats attacks. The Democrats, after all, want to spend much more than the Republicans are proposing. They would like to have higher taxes, but aren't willing to actually propose ending Bush's tax cuts, let alone offer new taxes, especially on their constituents.

The real problems is that the voters vote for deficits, and politicians know it. Voters like lower taxes and increased spending. Increasingly, the 90's are looking like a once-in-a-lifetime event to me. The bubble increased tax revenues; the Republicans were able to keep the Democrats from spending the increase, and the Democrats were able to keep the Republicans from cutting taxes. Deadlock worked to create surpluses while revenues were increasing. I don't see how surpluses could ever return with either party in firm control, or without unexpected tax revenue increases.

Posted by: Matt Judd on February 13, 2003 12:36 PM

Well, you did say that you agreed with Uncle Miltie's views on taxes and spending in a previous post, and offerred a qualified defense of Glenn Hubbard over on Brad's page.

Like I said, it's the larger context of how these arguments are used to justify the consistent fiscal irresponsibility of Republicans that rubs me the wrong way.

Can you deny that the Republicans have become the party if 'All Tax Cuts, No Sacrifice?'

Posted by: RC on February 13, 2003 12:52 PM

"if you look at late 90s/2000 budgeting, you can see that continued surpluses correlate strongly with massive increases in government spending (incluiding incresing entitlement benefits)"

Uh, no. Check out the table.

Posted by: Jason McCullough on February 13, 2003 12:54 PM

the effect of endogenous budget reductions under Clinton, rather than exogenous capital gains increases, are fairly trivial and certainly not what the Democrats like to claim in the interests of crediting Clinton for the economic boom.

I've never though the surplus had much to do with the boom. I'm sure it had some effect on growth rates, but not much. The good thing about the surplus, in my opinion, was that the government was sorta-kinda-hopefully attempting to get into good shape for when the boomers retire. Which, you know, is all gone now.

Posted by: Jason McCullough on February 13, 2003 12:58 PM

Matt Judd, I don't think the record bears out your argument of moral equivalency between Democrats and Republicans on the issue of deficits. Clinton and other Democrats before him were willing to take the politically difficult step of advocating tax increases. Bush isn't willing to cut spending.

Posted by: RC on February 13, 2003 01:05 PM

So both parties are practicing "Eat the Cake, too-ism" (a licensed phrase, so don't attempt to use it): the Dems want to increase spending but don't dare increasing taxes to keep the budget in balance; the GOP wants to decrease taxes but doesn't want to cut the programs to keep the budget in balance. So which approach is politically superior?

Posted by: JT on February 13, 2003 01:08 PM

Well, the Dems did increase taxes under Clinton, despite a lot of doomsaying from the Republicans, which proved totally inaccurate.

Bush 41 did too, of course, but got endless grief from the "taxes are the source of all evil" crowd in his party.

Posted by: Bernard Yomtov on February 13, 2003 01:34 PM

Matt is also wrong in that spending has risen faster under Bush than under Clinton.

So the GOP seems to be worse in all counts.

Posted by: GT on February 13, 2003 01:52 PM

Sure, but Bush 41 was operating in a recessionary period, whereas the Dems in the 90s were acting in a period of prosperity. People never like being parted from money, but under which circumstances will they protest more loudly?

Posted by: anony-mouse on February 13, 2003 01:59 PM

"Sure, but Bush 41 was operating in a recessionary period, whereas the Dems in the 90s were acting in a period of prosperity."

The Clinton tax increases occured in 1992, when there wasn't "prosperity."

Posted by: Jason McCullough on February 13, 2003 02:44 PM

Excuse me, 1993.

Posted by: Jason McCullough on February 13, 2003 02:47 PM

You guys are evading the question, which is about the future. I think I stated the real-life alternatives for voters in pretty clear terms. Which approach is likely to be more politically successful?

Posted by: JT on February 13, 2003 03:18 PM

The Democrats did raise taxes in 93, and have suffered politically for it. The Republicans tried to reduce government spending in 94, and paid a political cost for the shutdown. My point was that neither side now has any desire to limit deficits.

GT - I don't believe I suggested that spending has increased less under Bush than under Clinton. It doesn't surprise me a bit that it has gone up faster, since the Republicans in the House and Senate were the ones limiting spending under Clinton, and they are not willing to treat a President of their own party the way the treated Clinton.

I don't give either party credit for the surpluses of the late 90's. I didn't intend my comments to be taken as a defense of Bush; I agree with JT, that neither party will give us a balanced budget. I believe both would give us a large deficit; under Republicans, the government will both tax and spend less than under Democrats. Given those choices, I prefer the Republicans, but I could be won over by the Democrats if I believed they would actually balance the budget and spread the tax burden a little more evenly.

Posted by: Matt Judd on February 13, 2003 05:33 PM

My point was that neither side now has any desire to limit deficits.

What's the evidence that Democrats have no desire to limit deficits?

Posted by: Jason McCullough on February 13, 2003 09:23 PM

Jason: that Clinton had no interest in doing so until the Republicans wouldn't let him spend any money? Go find his proposed budgets and take a look at when he planned on getting rid of deficits. You'll see the major deficit reduction was supposed to kick in when he got out of office, a.k.a. never.

Posted by: Jane Galt on February 13, 2003 11:30 PM

Jason- The Democrats support larger increases in spending than do the Republicans on many issues, drug benefits for seniors being only one. They are as yet unwilling to attempt to roll back Bush's tax cut; they do not propose any increase in taxes, and in fact have their own (much smaller) tax cut plan for this year. Last year, while they controlled the Senate, they were unwilling even to publish their own budget.

What's the evidence that they are willing to limit deficits? Will they spend less than the Republicans, or increase revenues?

Posted by: Matt Judd on February 14, 2003 12:30 AM

It doesn't surprise me a bit that it has gone up faster, since the Republicans in the House and Senate were the ones limiting spending under Clinton,

We've been through this particular falsehood before.

Go check the numbers and you'll see you are wrong.

Spending began falling BEFORE the GOP took control of Congress.

Is there a reason why conservatives continue to push things we know to be false?

I mean even the Cato guys know this! They compare the first few years of Bush with the first few years of Clinton.

And, need I remind you that the first two years Clinton had a Dem Congress?

Posted by: GT on February 14, 2003 09:34 AM

Jane’s point that Clinton had no interest in reducing the deficit is also demonstrably wrong.

Clinton, BEGINNING WITH HIS VERY FIRST BUDGET, slowed the speed of government spending so that it fell as % of GDP while at the same time increasing revenues by raising taxes. This happened BEFORE the GOP took over Congress.

As a result the deficit began falling IMMEDIATELY and continued to do so until it became a surplus around 1998. The GOP taking over made no difference since they crashed and burned on the budget almost immediately (remember the disastrous govt shutdown) and more or less gave up after that. At best they simply continued what Clinton and the Dems began 2 years earlier.


Again, why do conservatives continue to push as facts things we know to be untrue?

Posted by: GT on February 14, 2003 09:54 AM

GT- When I said that the Republicans were restraining Clinton's spending, I was in fact talking about the years before Bush took over, when the Republicans were in control of the House and Senate. I do believe that Clinton deserves some credit for his first two budgets; he was a moderating influence on his party. Although as I remember it, the spending decreases came primarily from defense and the end of the S&L bailout, and the tax increase came on those who did not vote for him, so I temper my admiration. He did what came easily to a Democrat.

My point, though, was that the Democrats suffered politically for their virtue, just as the Republicans did after the shutdown. You are very good at defending the Democrats behavior ten years ago, which I'm not criticizing; you don't seem to be talking about their behavior recently, which I am criticizing. Do you disagree that the Democrats wanted to spend more money in the last few years under Clinton than did the Republicans, or that they want to spend more now than do the Republicans? Do you believe we would have any less of a deficit now if Gore had won and the Democrats were in control of the House and Senate? And could you stop criticizing us for things we didn't say and start responding to what we did say?

Posted by: Matt Judd on February 14, 2003 11:45 AM

Jason: that Clinton had no interest in doing so until the Republicans wouldn't let him spend any money? Go find his proposed budgets and take a look at when he planned on getting rid of deficits. You'll see the major deficit reduction was supposed to kick in when he got out of office, a.k.a. never.

Check out table 8.4. In 1993, total nondefense spending went up .1% of GDP, and in 1994, it went down .1% of GDP; those were the two years the democrats had completely unfettered control of the government.

In table 1.2, on-budget receipts increased by .1% of GDP in 1993 and .5% of GDP in 1994.

This somewhat lowballs the effect of the tax increase, as I think it was phased in over a few years.

The other area to look at is that maybe the democrats increase means-tested entitlements, like AFDC and food stamps; that BEA category containing those went up by .1% of GDP in both 1993 and 1994. Look at the line items that democrats could have increased in table 8.6:

1) Education, training, employment, and social services was 9.6 billion in 1992, 11.7 billion in 1993, and 6 billion in 1994 (constant dollars).
2) Food and nutirition assistance 32.2 billion in 1992, 33.8 billion in 1993, and 34.5 billion in 1994.
3) "Family and Other Support Assistance", which I think is AFDC and related: 16.5, 16.7, 17.2.

These sum up to -3.6 billion, +2.3 billion, and +.7 billion, or roughly zero; they certainly fell behind the rate of GDP growth.

That's pretty much it; the other means-tested entitlements are stuff like EIC and Medicare where there isn't much year-to-year control over allocations.

In summary: over the two years of democratic control, the democrats:

1) Did not change spending.
2) Increased revenue by 4.5% over the baseline.

Fits pretty well with "Clinton and the democrats focused on reducing the deficit from the start, albeit not very much." Does not fit well with "democrats really increased spending."

Posted by: Jason McCullough on February 14, 2003 11:50 AM

You can't complain about "the democrats want to spend more" while ignoring "the republicans want to tax less" when talking about the deficit.

Posted by: Jason McCullough on February 14, 2003 12:05 PM

"Again, why do conservatives continue to push as facts things we know to be untrue?"

Simple answer: Basic Republican principles of are:

1. Anything good that happened while Clinton was President was the result of the wisdom of Reagan and Bush Sr.

2. Anything bad that happened was Clinton's fault.

3. Anything good happening now is the result of Bush Jr.'s wisdom

4. Anything bad happening is either Clinton's fault or would have been worse under Gore.

As Calpundit has pointed out, NRO has recently attempted to stretch these principles even further, blaming Clinton for deficits incurred by Bush Sr.

Posted by: Bernard Yomtov on February 15, 2003 11:19 AM

The evidence is mixed on the original question. No breakthroughs there.

Its easier for politicians to create new programs then to cut existing programs. What gets cut in one year is replenished a year or two later.

Both parties mislead. Republicans should cut spending and change corresponding laws that discourage the private sector from providing whatever service the eliminated program did. Their presmise is that private actors do it better then public actors. Democrats should cut obsolete programs and replace them with ones that might work better. Their premise is that public actors do it better then private actors.

Posted by: Keith on February 17, 2003 11:33 AM

Comments are Closed.